THE RACE

The Race is about our standard of living and how we

can increase it. Today, we are facing a real threat in the
Western World that the opposite will happen.
This companion book to Eli Goldratt and Jeff Cox’s
underground bestseller, The Goal, is helping reverse this
situation.
These two books have given rise to Goal Circles and
Race Circles—excellent first steps in beginning a process
of ongoing improvement._____________________ ________

Eliyahu M. Goldratt
and
Robert E. Fox
THE RACE

By
Eliyahu M. Goldratt
and
Robert E. Fox

North River Press, Inc.

Copyright © 1986 Creative Output Inc.

AU rights reserved. Except for the usual review purposes, no part

of this work may be reproduced or transmitted in any form or by any
means, electronic or mechanical, including photocopy, recording, or
any information retrieval system, without the written permission of
the publisher.

Manufactured in the United States of America

Library of Congress Cataloging-in-Publication Data

Goldratt, Eliyahu M., 1948-

The race.

I.Inventory control. 2. Quality of products.

3. Corporate profits. 4. Efficiency, Industrial.
5. Competition, International. I. Fox, Robert E.,
1937- . II. Title.
HD40.G65 1986 658.7'87 86-18197
ISBN 0-88427-062-9

For more information on the concepts'presented in this book please contact:

Creative O utput
Commerce Park
M ilford, CT 06460
Tel: (203) 877-5671

For additional copies please contact your local bookstore, Creative Output, or the publisher:
North River Press
Box 241
Croton-on-Hudson, NY 10520
Tel: (914) 941-7175

4 5 6 7 8 9 10
 INTRODUCTION

THE RACE is about our standard of living and how we can increase it. Today, we are facing
a real threat in the Western World that the opposite will happen. The threat is not just to us.
If our living standard declines, everyone else on this globe is adversely affected.
Both the source and the solution to the problem lie in the same place—manufacturing.
Manufacturing has been the major wealth generator of our industrialized world. This ability
to generate wealth has made our standard of living the envy of the rest of the world. If we
continue to lose our manufacturing base, and we are losing it rapidly, we and everyone else
will certainly live less well.
This companion book to Eli Goldratt’s and Jeff Cox’s underground best seller, THE GOAL,
can help you to reverse this situation. Readers of THE GOAL, and they range from the
boardroom to the universities to the factory floor to spouses at all levels, report startling
improvements just because their company read a business book disguised as a novel, or as
some say, a love story.
THE RACE is also an unusual book. Its messages can be grasped simply by looking through
the graphics. It can be understood better by reading the accompanying text. It is even more
deeply grasped and useful when manufacturing people at all levels discuss its implications and
application to their own environment.
The introduction of THE GOAL caused a development we did not expect—GOAL Circles.
Action groups of both managers and shop floor employees working to find their ‘‘Herbies,”
cut batch sizes, abandon efficiencies and link marketing actions more closely to manufacturing
capabilities. Distribution of some galley proofs of THE RACE seems to be causing the same
phenomenon—RACE Circles. Maybe the Western World is developing its own response to
the Japanese Quality Circles.
GOAL Circles and RACE Circles are an excellent beginning. Through them companies can
begin a process of ongoing improvement. However, one-time improvements, no matter how
significant, will not allow us to catch up or stay ahead for long in the increasingly intense race
for a competitive edge in which our industries are embroiled. We need to permanently establish
and continually improve on a process of ongoing improvement.
THE RACE enables you to derive a superior system— Drum-Buffer-Rope— for generating
continual logistical improvements. It also illustrates how to focus on the process improvements
that will have the greatest impact on your competitive edge. The epilogue and appendix quizzes
will give the thoughtful reader insight in how to initiate and then extend a process of ongoing
improvement into other areas like marketing and financial control.
Each step in implementing an increasingly effective process of ongoing improvement will
require better understanding and more effective tools. A company often needs to look outside
for the insights and products to establish such an approach. We believe, however, that the
revolution required in how you run your companies can only be led from the inside. The steps
required to adopt and manage a process of ongoing improvement are a management respon­
sibility.
The reversal of the fortunes of our manufacturing companies is certainly important to its
employees and investors. It may be even more important to each of us and our countries. We
wish you much success and hope THE GOAL and THE RACE help to move you in the correct
direction.

Eli Goldratt
Bob Fox
 1: ARE WE LOSING OUR DOMINANT
POSITION?
The Industrial Revolution began in England and spread across Europe and America
in the ensuing years. Since that beginning, Western countries have held a predominant
position in almost all types o f manufacturing. The strength of these industries made
our standard o f living the envy o f the world.
In the last 15 years a dramatic change has occurred. In 1970, it became apparent
that we had lost our predominant position in smokestack industries like steel, brass
and textiles. Our world market share slipped and plants were closed due to competition
from the East. We thought our demise was due to the low-cost labor and modem
equipment of our competitors.
Five years later, our dominance in electrical appliances disappeared as stereos,
televisions, microwave ovens and other goods flooded in from Japan and other Far
Eastern countries. This time we blamed our losses on dumping and copying o f our
products.
In 1980, when the pride o f our manufacturing prowess, the automotive industry,
was clearly in jeopardy, we began to realize that the problem was truly serious. W hile
we found additional reasons to explain our problems, concern about our competitiveness
was now becoming widespread.
We thought we still had security in high technology and new product development.
Now we know that even these niches are not safe. In 1985 we lost our leadership in
the production of microchips, the very heart o f the information age. Signs now indicate
that the aerospace industry will be the next place where we lose our leadership position.
Even the warm security blanket o f the Department of Defense will probably be insuf­
ficient to protect this industry beyond 1990.
In 15 short years we have seen an unprecedented change in a host of industries from
smokestack to high tech. The ramifications for our economic well-being and standard
of living have just started to be felt. It's time we realized that this sharp shift is not
the result o f trivial causes or the excuses we have used. It’s because of an unprecedented
race in all aspects o f manufacturing. L et’s examine just a few to understand the
accelerating pace o f the race for a competitive edge.

2
ARE WE LOSING THE RACE?
# 1970—Smokestack

0 1975—Appliance

# 1980—Automotive

t 1985—Electronics

# 1990 (?)—Aerospace
 2: THE DRIVE FOR QUALITY
Quality is probably the best aspect o f the race that enables us to understand its impact
on the marketplace. Until 1970, we used the word yield as the measure o f quality. Our
concentration was on how many good parts resulted from the material we put into the
plant. This terminology indicates that probably more than 10% o f the parts were
scrapped. Now we use the word yield only when referring to the start-up o f new
processes and products where we know that initial losses will be high.
During the 70s our terminology changed. We adopted the word scrap to denote the
shift in our focus from good parts to defective ones. Improvement occurred and our
losses dropped below 10%. In 1980 we realized that even this improved level o f quality
was insufficient. We began to lose markets because our competitors supplied products
o f superior quality. The Japanese penetration o f the American automotive industry is
a classic example. Consumers were initially attracted by the promise o f low-cost,
economical transportation. Once they became accustomed to the quality and depend­
ability of Japanese cars, a new standard was set in the marketplace. The American
automotive companies lost market share because o f lower quality and soon realized
that improving it was essential if they wanted to remain in business. Slogans like
“ Quality is number one” and “ Quality is job on e” appeared on the scene, signifying
the drive to reduce scrap below 1%.
Now our terminology has changed once again to reflect the even higher level o f
quality that the market is demanding. We have started to talk again in terms o f “ zero
defects. ” The magnitude o f this change is seen in the new unit o f measurement
introduced by the Japanese to gauge progress toward this goal— parts per million scrap.
Parts per million means that our quality goal has increased four orders of magnitude
in the last 15 years. We are seeing an unprecedented acceleration in the race for higher
quality. But quality is not the only aspect o f the race.

4
QUALITY
• Before 1970-Y ield (Defects: ~ 10%)

# 1970—1980—Scrap (Defects: < 10%)

t 1980—“ Quality is number 1”
(Defects: < 1%)
# 1985 —Zero Defects
(Defects: parts per million)
 3: PRODUCT LIFE CYCLES SHRINK
Consumers have not only been insisting on higher quality, but they are also vora­
ciously consuming new products. Prior to 1970, we were accustomed to buying products
that changed only slightly each decade. As we moved into the 70s, new products
entered and were accepted in the marketplace at an increasingly faster rate.
We experienced the dizzying pace o f change as electronics were introduced into
previously staid products like watches and calculators. We moved from an era where
a good watch not only lasted a lifetime, and was even passed on to the next generation,
to a disposable world where we purchase new watches with every improvement in
technology or change in fashion. Calculators changed from bulky mechanical and
electrical machines with limited capability to miniaturized electronic marvels rivaling
earlier computers. Instead o f being limited to offices, calculators are now found in
almost every purse, billfold and schoolbag. The engineer’s slide rule has become a
museum piece.
The shortening o f product life cycles was not confined just to consum er products.
Every industrial organization has been impacted. Even a time-honored product like the
2 x 4 is being replaced by galvanized steel structurais in many buildings. The pressure
for new products could not be supported by our conventional design methods. We had
to develop computer-aided design and computer-aided engineering systems to cope
with an ever-increasing pressure for new products. By 1980, the demands of the market
and increased capabilities o f our design processes shortened life cycles to just a few
years.
Today we are in a period where, if a company is late to the marketplace with a new
product by six or nine months, it runs the risk o f losing the entire market. Certainly
the life cycle o f most products has shrunk again. We can only guess the life cycle of
products today. We will know tomorrow.

6
PRODUCT LIFE CYCLES
# 1970—Decade(s)

t 1975—Many Years

t 1980—A Few Years

• 1985—?
 4: AUTOMATING OUR PLANTS
This race for a competitive edge is also manifested in ways less obvious to the
typical consumer. Machine technology has been changing at an explosive rate. Prior
to the 70s, we used conventional electromechanical equipment that had changed little
in the last 40 or 50 years. By 1975 we had introduced computer technology into
production operations in the form o f numerically-controlled (NC) equipment.
At that time we might have logically projected that the adoption o f this new tech­
nology would take decades, since it involved very large investments and a substantia]
retraining of the work force. Remember, it was the first time we introduced computers
and programming onto the shop floor. Nevertheless, this technology is commonplace
today.
By 1980, even before NC technology was widespread, we had introduced the next
generations— CNC (Computer Numerically Controlled) and DNC (Direct Numerically
Controlled) equipment. Instead o f stand-alone NC machines, we now had cells, or
groups o f machines, interlinked and controlled from a single computer source. Despite
this rapid change, by 1985 we were forced to invest additional huge sums to follow
the Japanese in the introduction o f Flexible M anufacturing Systems (FMS) . Changing
markets and management perceptions demanded flexible computer-controlled equip­
ment capable of handling small-lot production, and rapidly changing product designs.
Even while we are still investing in FMS we can see the next step around the com er.
Many major manufacturers are already investing tens o f millions o f dollars in the
attempt to build totally automated factories, the lightless plants, a technology which
even today is unclear in its design and usage.
Everyone is taking tremendous risks in the attempt to participate and not fall behind
in this race. The pressures to gain a competitive edge through machine technology are
great, but they are even more intense and obvious in the frantic attempts to find a better
logistical system to run our plants.

8
MACHINE TECHNOLOGY
# 1970—Conventional Machines

f 1975—Numerically Controlled
(NC) Machines
# 1980—Cells of Computerized
(NC) Machines

t 1985—Semi-Automated Plants
t 1990 (?)—Lightless Plants
 5: LOGISTICS TO SPEED THE FLOW OF
MATERIALS
During the 40s, 50s, and into the 60s, we used manual “ order point” techniques
to control -the ordering and flow o f material in our plants and warehouses. This was
our logistical system. About 1965, we tried for the first time to tap the power o f the
computer for this task through a technique called Materials Requirements Planning
(MRP) . Despite an investment estimated at $10 billion, we were not satisfied with
the results. In 1975 we renamed it closed loop MRP, believing that feedback on the
status of shop orders and purchase orders was the key to faster material flow. In 1980,
it was MRP II, an effort to get the entire manufacturing organization— marketing,
engineering, manufacturing and finance— singing off the same hymn sheet.
Each phase o f our MRP journey involved large investments in computers, software
and training in how we managed our businesses. It’s been estimated that w e’ve spent
more than $30 billion, but even these enhancements and investments were not enough.
MRP did not enable us to maintain leadership in the race for a competitive edge. The
Japanese approach to the logistics o f the shop floor, Just-in-Tim e/Kanban, proved
superior to our efforts. Today some W estern companies are attempting to emulate
them. Meanwhile the Japanese and others are searching frantically for an even better
system, called synchronized manufacturing, even though we have yet to define exactly
what it is.

10
LOGISTICAL SYSTEMS
0 1950—O rder Point
• 1965—MRP

t 1975—Closed Loop MRP

• 1980—MRP II

0 1985—Synchronized
M anufacturing

11
 6: INVENTORY TURNS MEASURE
PERFORMANCE
Nowhere has the race manifested itself more than in inventory turns. The rate o f
turnover or usage o f inventory is an excellent measure o f the performance and rate of
change o f manufacturing companies. During the 70s the acceptable standard for in­
ventory turnover was somewhere between two and five a year. A study by the inter­
national consulting firm o f Booz, Allen & Hamilton showed that U. S. firms averaged
3.7 turns during the 70s. The Japanese average, while higher, was still only 5.5.
These were the standards. Those with the foresight and courage to suggest that double­
digit turns were possible were labeled as lunatics.
In the current decade, two to five inventory turns is considered totally inadequate.
The standard o f acceptable performance has shifted dramatically in just a few years
to somewhere between five and 20. Inventory turns o f two digits that were considered
to be impossible just a few years ago are now a must. A number o f western companies
already are operating in the range o f 30 to 80 inventory turns. Some Japanese companies
(thank heaven there are only a few) have demonstrated that inventory turns o f three
digits can be achieved. Everywhere there is a drive to do much, much better than was
previously thought possible.
Even with these dramatic new, and as o f yet generally unreached goals, another new
target is emerging on the horizon— negative inventory turns. Turning inventory so
rapidly that we are paid for the finished product before we have to pay for the raw
materials is now considered a possibility. W e’re probably already paying for the ham ­
burger we buy at a fast food store before they have to pay for the meat. Maybe it’s
also possible in a manufacturing company. It’s quite a change when inventory is viewed
as a source rather than as a user o f money.

12
INVENTORY TURNS
1 Before 1980 — 2-5

• After 1980 - 5-20

# 1985 — Some already at 30-80

• 1985 (JAPAN) - A few have

demonstrated > 100
# Tomorrow (?) — Negative Inventory
Turns

13
 7: THE RACE FOR A COMPETITIVE EDGE IS
ACCELERATING— RAPIDLY
These few examples illustrate that the race for the competitive edge is clearly in­
tensifying. The challenge that the W estern manager faces is how to become much more
competitive very quickly. W e have entered a period unlike any since the days o f the
industrial revolution. The implications for companies, countries and our standard o f
living are every bit as profound as when the first industrial companies began to appear
in England years ago.
It is no longer a question o f a cycle o f good times and bad. We cannot batten down
the hatches and hope to survive, as if this were another passing storm. We can no
longer use the conventional approach o f cutting expenses and firing people in the bad
times. We must find a way to continually improve— in good times and bad. We must
choose to be in the competitive edge race. The companies that elect to shrink in order
to pass the bad times will simply disappear. The ones that survive will be those
companies that will find a way to participate in this ever-increasing competitive edge
race.
The magnitude o f our problem is revealed once we accept that we are behind and
recognize that we have only a little time left. O ur resources, particularly management,
are severely limited, and our funds are not sufficient to allow us risky experiments.
We must make the correct decisions this time or be prepared to live with the conse­
quences.

14
Race Intensity

1970 1980 1986

How to become more competitive considering

# Limited time

t Limited resources
# Limited funds

IS
 8: WHERE TO START—SORTING OUT THE
ALPHABET SOUP
In the last five years, a host o f new and powerful techniques have emerged as
possible solutions. A whole lexicon o f new terms has been developed that seems to
be characterized by two- and three-letter abbreviations. We are urged to implement
statistical process control (S P C ), group technology (G T ), factories o f the future (FOF)
and on and on. The W estern manager is challenged to solve a very fundamental problem
from this alphabet soup o f solutions.
To understand each o f these new technologies can, by itself, be a time-consuming
challenge. Deciding which are best is a formidable task. Figuring out how to put them
all together seems beyond our reach. Since we don’t have the time, resources or funds
to do everything, everywhere, we had better be convinced that we are taking the actions
that will leapfrog us back into the race. There is no longer margin for error and no
time for risky experiments.
What are the most important steps to take first and which technologies will give us
the greatest return so we have the time and funds to implement the rest? Maybe we
should begin by reexamining the goal o f this race. By coming to grips with the basic
purpose and nature of our manufacturing organizations maybe we can find a path that
will enable us both to participate in and win the race for a competitive edge.

16
 % *o.

G ^ Y
&

^
A ^°V ^
~ ,
X >
4$r % '% ,
nd ^ % % ,
<*-° % '^ v
-v x 1b ^
v -c ® * \

THE COMPETITIVE EDGE \

RACE:
WHERE DO WE START?

17
 9: WHAT IS THE GOAL OF THE RACE?
What are we really trying to accomplish in our businesses? Did the investors and
employees put their money and efforts into the company with the altruistic intent of
providing-better service to their customers? O r did they do it because they wanted the
prestige o f having the largest market share? Did they want to boast that they had lower
costs than their competitors? Was the goal o f these investors and employees to enjoy
the pride o f having the highest quality products? Is it likely that they invested their
money and efforts to build the company just in order to have it survive?
We think not. All of these targets may be useful means to the goal o f the company,
but they themselves are not the goal. We believe that the goal of a manufacturing
company is one and only one— to make money in the present as well as in the future.
That is what winning the race is all about. But what does “ making m oney” really
mean?

18
WHAT IS THE GOAL?
Better Customer Service?
Larger Market Share?
Lower Cost?
High Quality?
Survival?
♦
Make MONEY in the present
as well as in the future
*
Win the race

19
 IO: “ MAKING MONEY”—HOW DO WE
MEASURE IT?
We all know the bottom line measurements o f making money. A company needs
to make a net profit, an absolute measurement o f making money. But is this meas­
urement by itself sufficient? If a company made $10 million, is that good or bad? If
they had invested $20 million, it's quite good. But if the investment were $200 million,
it’s lousy. We need an additional measurement that shows how much money we made
relative to the money we invested in the business, a measure like return on investment.
These two measurements seem sufficient, but many a company has been rudely re­
minded by the threat o f bankruptcy, that there is also a survival measurement, like
cash flow. Cash flow is an on-off measurement. When we have enough cash, it is not
important. When we don’t have enough cash, nothing else is important.
While these three bottom line measurements are sufficient to determine when the
business is making money, they are woefully inadequate to judge the impact of specific
actions on our goal. For example, in what size batches should we process material
through our plants? Five? Fifty? Maybe five hundred? How will these batch sizes
impact the bottom line o f the entire company? Or, should we buy a new robot? It
certainly is going to be more efficient, but it is also expensive. W hat will be the
resulting impact on our financial measures? O r should we accept an order for a product
when the selling price is substantially below our standard price? We clearly need some
type o f bridge between the specific operational decisions we must make and the bottom
line measurements o f the entire company.
Today, our bridge is based on the cost concept. We have developed a host of
procedures and systems based on the idea o f cost. We employ the technique o f economic
order quantities to help determine batch sizes. W e analyze investment opportunities
based on cost reduction to determine where to employ our capital. We calculate product
costs and margins to help understand which products we should push in the market
and which we should discontinue.
The cost concept and cost procedures are the current bridge between our actions and
the bottom line measurements, but is this bridge taking us in the right direction?

20
THE GOAL:
TO MAKE MONEY

Bottom line measurements

RETURN ON
NET PROFIT INVESTMENT CASH FLOW
(Absolute) (Relative) . (Survival)
\ _____________________

[What is the bridgeF )

ACTIONS

21
 11: IS THE COST BRIDGE FATALLY
FLAWED?
Despite the fact that cost procedures are well defined, managers commonly override
the calculated results because o f their experience and intuition. Economic order quan­
tities are ignored. Instead o f using the calculated lot size o f 46. 5, we chose 50 and
then override this decision when we split and overlap batches in our plants. M ajor
investments are currently being made in factories o f the future where the projected
return on investment is seven or eight years even though our required return for other
investments is less than three years. We decide that factories o f the future are strateg­
ically important regardless o f what the cost calculations show. Orders are accepted by
plants even though they shouldn't be according to cost recommendations.
Currently we are using both cost and intuition to determine what actions to take.
The mere fact that we override the cost recom mendations so frequently already tells
us that cost procedures are not adequate. Intuition often helps to improve the cost
recommendation, but unfortunately intuition is not a basis for good communication.
Even though this combination o f cost and intuition were not sufficient in the past, this
new competitive race has now made them totally obsolete, as can be seen by the next
two examples.

22
THE BRIDGE:
Currently we are using cost (+ intuition).

Is it good enough in this new

competitive world?

23
 12: THE COST CONCEPT BLOCKS BETTER
QUALITY
The competitive edge race is forcing us to reduce scrap to improve quality. Suppose
a company saves $10,000 a year in labor and material every time it reduces scrap by
1%. If the scrap level was 5% and the company was able to double the quality level
by reducing scrap to approximately 2% , the savings would be $30,000 a year. L et’s
assume that the investment in equipm ent, tools or education required to achieve this
improvement was $20,000. The return on this investment would be less than one year
and the cost judgm ent would be to do it.
L et’s see how the use o f this cost bridge impacts the ever-increasing steps to improve
quality in the race for “ zero defects. ” Doubling quality again by further reducing
scrap from 2% to 1% would result in a savings o f only $10,000. The investment
required now is undoubtedly higher than $20,000 since the first step probably required
resolution of only one or two major problems while this step may involve several
investments to resolve many smaller problems. Nevertheless, let’s assume that the
investment is still only $20,000. The decision now is less clear since two years are
required to recover our investment.
The savings from a third doubling o f quality by a scrap reduction from 1% to 0. 5%
will only be $5,000. O ur investment most certainly will be much larger than $20,
000, but let’s be very conservative and again use this amount. Now the cost judgm ent
will clearly be don’t do it since the four-year return does not justify the investment.
How can we hope to reach a quality level o f parts per million scrap when the cost
bridge already blocks us around the 1% level? W e know that when we produce defective
products we don’t just scrap material and labor, we are also scrapping our market. In
our increasingly competitive world, we are forced to declare that “ quality is number
one, ’’ which means that every investment that involves quality improvement is allowed.
We have totally lost the yardstick. The bridge in this area simply no longer exists.
The same cost judgm ent phenomenon that helped us in the past, and now blocks us
from participating in the race, is also revealed in the frantic attempts to reduce inventory.

24
 The cost concept blocks
us from achieving a competitive edge
through parts-per-m illion quality.

Reducing
Scrap Annual Investment The Cost
From To Cost Savings Needed Judgment

5% 2% $30,000 $20,000 Do

2% 1% $10,000 $20,000 Indifferent

1% 0.5% $5,000 $20,000 Do NOT Do

25
 13: HIGH INVENTORY TURNS AREN’T COST
JUSTIFIED
Suppose a company has an inventory o f $ 15 million and estimates the cost of carrying
inventory at about 25% o f its value. If the company is currently turning its inventory
three times a year and is able to double the turnover to six, costs will be reduced by
approximately $2 million a year (25% x $15 million/2) . Such an inventory reduction
is not achieved without efforts and investment. If the investment needed is around
$2 million, then the cost judgm ent will be to do it since the return is approximately
one year.
The situation changes when the company wishes to double inventory turns once
again from six to twelve. Now the cost savings are only $1 million (25% x $15
million/4) . The investment required is probably higher. N evertheless, even if we use
the same $2 million, the decision now is unclear because the payback is over two
years. The cost judgm ent is neutral. If there are some good intangible reasons for
investing, do it. If there aren’t, d on’t. But when the company is taking the necessary
step for survival in this new competitive race and attempts to double inventory turns
from 12 to 24, it is blocked. The cost savings are now only $500,000, a four-year
return. Now the cost judgm ent is loud and clear. D on’t do it.
It is very obvious that we must look for a better bridge to guide us in our efforts
to catch up. Fortunately, there is a widely used set o f three measurements that are not
bottom line measurements and not cost measurements. These measurements are total
sales, total inventory and total operating expenses. The use o f these measurements is
widespread because our intuition is already telling us that they are a sound bridge.
However, to use this throughput-inventory-operating expense bridge for specific de­
cisions we need to develop procedures to guide our actions. As a first step in this
process we should define precisely what we mean by throughput, inventory and op­
erating expense.

26
 The cost concept blocks
us from achieving competitive edge
through high inventory turns.

Increasing
Inventory Cost Savings Investment The Cost
Turns Per Year* Needed Judgment

3 6 $2M $2M Do

6 12 $1M $2M Indifferent

12 24 $0.5M $2M Do NOT Do

*Assuming starting inventory of $15M and 25% carrying cost

27
 14: THROUGHPUT-INVENTORY-OPERATING
EXPENSE—A BETTER BRIDGE
We have chosen to define these three measurements in the following way:
Throughput - The rate at which an organization generates money through sales.
Notice, through sales, not through production. If we have produced something and
haven’t sold it, it’s not throughput.
Inventory - All the money that the system invests in purchasing things that it intends
to sell. This definition o f inventory deviates from traditional definitions since it excludes
the added value of labor and overhead. We have elected to use this definition in order
to eliminate the distortions and counter-productive decisions caused by accounting­
generated inventory profits and inventory losses.
Operating Expense - All the money the system spends in order to turn inventory into
throughput. This definition o f operating expense includes not just direct labor, but also
management, computers, and even the secretaries. If a secretary’s jo b is not to assist
in turning inventory into throughput, then her salary is not a real operating expense,
but just a waste.
A much broader explanation o f the meaning o f these measurements can be found
in The Goal - A Process o f Ongoing Improvement. We call these three measurements
the global operational measurements. The first step in using them as a bridge is to
clarify the linkages between these operational measurements and the bottom line fi­
nancial measurements.

28
THE COST CONCEPT MUST
BE REPLACED BY GLOBAL
OPERATIONAL MEASURES.

TH RO UG H PUT-
77^ rate at which the system
generates
money through sales.
INVENTORY—
All the money the system invests
in purchasing
things the system intends to sell.
OPERATING EXPENSE—
All the money the system spends
in turning
inventory into throughput.

29
 15: T-I-OE AND THE BOTTOM LINE
We know that our goal is to make money and that we measure progress toward it
by the three bottom line financial measurements. If we take actions that increase these
three measures simultaneously we are certainly moving in the right direction.
The intuitively felt connection between throughput-inventory-operating expense (T-
I-OE) and the bottom line measures are sharpened with our new definitions. Now we
can see that when throughput is increased without adversely effecting inventory and
operating expense, then net profit, return on investment and cash flow are simulta­
neously increased. We achieve the same result when operating expense is decreased
without an adverse affect on throughput or inventory.
When we analyze the impact o f reducing inventory we see that the result is not the
same. Decreasing inventory directly increases only return on investment and cash flow.
It does not have any direct impact on net profit. Should we conclude that inventory
is less important than throughput and operating expense? That seems to be the way we
have regarded it.
It’s also the way most managers have historically viewed T-l-O E. Net sales (through­
put) and total operating expense have always been seen to be important. Inventory has
frequently taken a back seat. When we look more closely, we see that inventory does
impact net profit and also has an additional effect on the other two bottom line meas­
urements. However, these impacts are indirect— through the carrying charge channel.

30
THE DIRECT IMPACT:
OPERATIONAL
MEASUREMENTS AND THE
BOTTOM LINE

THROUGHPUT INVENTORY OPERATING

* EXPENSE

/ v \

31
 16: INVENTORY AND THE CARRYING
CHARGE CHANNEL
The indirect impact of inventory on the three bottom line measurements is typically
estimated through the use o f carrying charges. We recognize that lowering inventory
reduces a number o f operating expenses, such as interest charges, storage space, scrap,
obsolescence, material handling and rework. Recently most companies have been
estimating their annual carrying charges to be some number around 25% of the value
o f the inventory (value including labor and o v erh ead ). Since reducing inventory lowers
operating expense, it increases the three bottom line measurements.
We can now see that reducing inventory has a twofold impact on return on investment
and cash flow due to its direct and indirect impact. There is only one indirect impact
on net profit and that is through the carrying charge channel. This is how W estern
management and our financial systems have generally regarded inventory. We saw in
an earlier example how our current financial methods lead us to believe that as we
lower inventory, the importance o f savings from further reductions of inventory di­
minishes. This is in drastic contrast to the approach to inventory that our worthy
competitors, the Japanese, are taking.

32
THE INDIRECT IMPACT:
INVENTORY AND CARRYING
CHARGES-
THE WESTERN VIEW
RETURN ON
NET PROFIT INVESTMENT CASH FLOW

33
 17: WHO IS RIGHT—OUR FINANCIAL
SYSTEMS OR THE JAPANESE?
The Japanese place enormous emphasis on reducing inventory. In fact, they go to
such an extreme that they characterize inventory as “ an evil. ” They make tremendous
efforts to reduce inventory no matter how small it is already. The Japanese approach
has recently been advocated in the USA by the American Production and Inventory
Control Society (APICS) under the name o f “ zero inventory” to illustrate this all-
consuming drive to eliminate inventory.
If you had to decide which is right, our financial systems or the Japanese, what
would be your choice? W e, like you, believe the Japanese probably are right. We
believe that there is a second indirect impact o f inventory, one that is not yet recognized
by our financial systems.
There is a growing awareness in W estern industry that there are “ intangible” benefits
from reducing inventory, which in this frantic race are more important than anything
else. We are increasingly recognizing that inventory has significant impact on our
competitive position in the market. In order to clarify this impact, we should examine
the elements that compromise competitiveness in the market.

34
The traditionally recognized impact of
inventory improvements diminishes rapidly as
inventory decreases.

So why do the Japanese

place so much emphasis on inventory
reduction?

The Japanese approach is advocated in the

USA under the name
“ ZER O -IN V EN TO R Y . ’ ’

35
 18: THE SIX COMPETITIVE EDGE ISSUES
We can gain a competitive edge by having better products, lower prices or faster
response. Oddly enough, each o f these categories can be separated into two distinct
branches. A competitive edge can be gained through our products by both superb
quality and excellent engineering. For exam ple, suppose that two companies are offering
the market the same product for the same price, but one o f them produces a much
better quality product. Certainly, the company with superior quality will eventually
capture the market. We are already well aware o f one startling example— Japanese
cars. The Japanese have penetrated W estern markets because o f their superb quality
and not because o f more features or faster delivery. On the other hand, we can gain
a competitive edge because our products are better engineered than our competitors.
If two companies offer the market the same type o f product at the same price and
quality, then the company that offers more product features will certainly capture the
market.
The same pattern holds true where price is concerned. The company with the highest
margins (lowest cost) will have more flexibility on pricing and thus can capture the
market. But we shouldn’t ignore the tremendous advantage in lower investment per
unit. This competitive edge also gives a company greater flexibility to compete because
of its lower break-even point.
Responsiveness is likewise composed o f two parts. The first is the competitive
advantage stemming from better due-date performance. We promise to deliver a certain
quantity o f product by a certain date. How many times do we succeed in fulfilling this
commitment? If we accomplish it 80% o f the time while our com petitor’s performance
is consistently 90% , he will eventually get our customers. If we are delivering as
promised 90% of the time and our com petitor’s performance is 95% , he will still win.
It is a never-ending race since customers are constantly increasing their expectation
levels, always adjusting to the high performer as the standard.
Due-date performance differs from shorter-quoted lead tim es, the second respon­
siveness avenue. This advantage gives us the ability to commit to an earlier delivery
than our competitors. Every salesman has learned the significant advantage in offering
the client fast deliveries.
We believe that these six elements comprise the competitive edge issues in today’s
and tom orrow’s market. The real race today is not just on one o f them , but on all six.
Oddly enough, most o f these elements are considered by our financial systems as
intangibles. Maybe they should be thought o f instead as our future throughput. We are
going to show, through a simple exam ple, the real impact o f inventory on our future
throughput and the six competitive edge elements.

36
A ROLE FOR REDUCED
INVENTORY?
QUALITY

PRODUCT
ENGINEERING

HIGHER MARGINS

PRICE.
LOWER INVESTMENT PER UNIT

DUE-DATE PERFORMANCE

RESPONSIVENESS
SHORTER QUOTED LEAD TIME

37
 19: A HIGH INVENTORY ENVIRONMENT
We can explore the impact inventory has on the six competitive edge elements by
contrasting a high inventory manufacturing environment with a low inventory one.
Suppose a company had an order for 1,000 units which are manufactured in a five-
step process. In high inventory manufacturing, the material might be processed and
moved through the plant in a single batch o f 1,000 pieces. Each operation completes
all its work before any o f the material is moved to the next operation. As material is
released into the plant, the work-in-process inventory level rises and does not begin
to decline until the product is completed at the last operation and can be shipped.
In this high inventory example, it takes about four months to complete the order
even if we run the plant 24 hours a day, seven days a week. This is in marked contrast
to low inventory manufacturing as we shall see.

38
HIGH-INVENTORY
MANUFACTURING
4 MONTHS ---------------- > ORDER
1000
UNITS

INVENTORY

AVERAGE
INVENTORY

HOURS 1000 2000

39
 20: A LOW INVENTORY ENVIRONMENT
In our low inventory manufacturing example, there are only two changes. First we
have split and overlapped the batches. We no longer wait until each operation has
completed the entire order before moving completed parts to the next operation. Material
is moved between operations in batches o f less than 1,000 pieces allowing several
operations to work on the same order simultaneously. In addition we have recognized
that in any process there is one operation that is the constraint, the operation more
heavily loaded or that takes more time than the others. In our example it is operation
C. Since we have recognized that C is the constraint, we have elected to release raw
material into the process only to keep the constraint, and not the first operation, busy.
As a result o f these two changes, the work*in-process inventory level is much lower,
and the order is completed in about half the time. While these benefits are attractive,
our real mission is to explore how our high and low inventory operations impact the
six competitive edge elements. It’s important to note that we are contrasting relative
environments, not absolute ones. The issue is not how much inventory a company has,
but how much it has relative to its competitors. The first competitive edge element we
will explore is quality.

40
LOW-INVENTORY
MANUFACTURING
ORDER
< ------- 2 M onths------ ► 1000
À UNITS

I I
HOURS 1000 2000

41
 21: DEMING TAUGHT THE JAPANESE
QUALITY
The Japanese have made a historic turnaround in the quality o f their products. In
the period after World W ar II we generally viewed Japanese products as being shoddy
and of poor workmanship. In just a few short decades, they have totally transformed
this image. Today the Japanese are setting the standard for product quality in many
industries.
How did they achieve such a remarkable change in so brief a time? If you inquire
o f the Japanese, they almost universally give the credit to Dr. W . Edward Deming,
an American statistician. With apologies to Dr. Deming for trying to articulate his
enormous contribution in a few words, his approach might well be summarized in the
statement that “ quality control should be used to check the process not the product!’
The secret o f Dr. Deming’s message lies not in the techniques that he taught the
Japanese, but in how he changed their focus and their attitude about quality. An example
o f the change in focus can be seen when a defective part is uncovered. Management
has two options. They can expedite a replacement part, or they can take the time to
determine the cause o f the problem. With limited management resources, it is difficult
to accomplish both tasks. The Japanese have elected to adopt Dr. Deming’s theory and
have focused their efforts on finding the cause o f the problem so they can eliminate
it forever. We continue to expedite.
Adopting Dem ing’s approach also requires a seemingly radical change in attitudes
about scrap and rework. So often the reaction o f W estern managers is to try to assess
blame (we also call it responsibility) . W ho was the careless worker, why w asn’t the
foreman paying closer attention, or who was the engineer that prescribed a faulty
process? As long as this attitude exists, it will be extremely difficult to find the cause.
O ur people will cover up the cause rather than expose it.
Adopting the Deming approach requires treating a defect not as a problem for which
someone should be blamed. Deming'believes that defects should be viewed as precious
jew els because they can help us find shortcomings in the process. If we can locate a
shortcoming, we have a chance to correct it for all times— to permanently improve our
process.
These are clearly two different approaches to quality, but what, if any, is the
relationship o f inventory to quality?

42
 Quality

Product

Quality control should check the

process, not the product.

—Dr. Deming

43
 22: LOW INVENTORY EQUALS HIGH
QUALITY
Suppose in manufacturing the 1,000 -piece order, that the product is damaged at the
first operation o f our process? This defect will eventually be caught, but where? Where
do we typically inspect our product? Unfortunately, often after the last operation. In
the high inventory environment, the damage will have occurred two months earlier,
making it very difficult to determine what caused the defect. W ho can recall what the
operating problems were two months ago? Even more importantly, we have great
pressure to expedite additional parts because the order is now going to be very late.
W here do you think management will devote their efforts, to expediting or finding and
solving the problem?
In the low inventory environment when the damage is detected at the last operation,
we are still producing the product at the first operation. We should be able to much
more easily determine the cause o f the problem. The pressure to devote extensive
management time to expediting now almost disappears. We have detected the problem
before the entire order was produced incorrectly. Fewer replacement parts are required
and we can produce them much more quickly than in the high inventory environment
even without resorting to expediting.
Management now has the time and ability to find and eliminate the cause o f the
problem— hopefully forever. It is probably not possible to have very high quality unless
we have low inventories. Certainly, there is a clear correlation between the people who
have the highest quality— the Japanese— and the people with the lowest inventory, also
the Japanese. Do you think that this is just a coincidence?
We can see the link between inventory and quality. Does a similar one exist between
inventory and product engineering?

44
HIGH VS. LOW INVENTORY
SYSTEMS:
OVALITY CONTROL
HIGH LOW
INVENTORY INVENTORY
Damage detected
*

after two months! while the damaging

What is the chance of process is still operating
locating the cause?

45
 23: PRODUCT ENGINEERING AND
INVENTORY—WHAT’S THE RELATIONSHIP?
The purpose o f product engineering changes is to improve our products, to make
them superior to those o f our competitors. If we can offer products that have the latest
functions'and features desired by the marketplace we can gain a competitive edge. The
power of new and better-engineered products is manifested in Wall Street’s fascination
with high tech companies. It is not the technology per se that draws investors but the
potential o f companies to be the first in the market with a new or improved product.
If we can be the first out with a faster, cheaper personal computer, or a new biogenetic
product, the world is our oyster.
Almost no sector o f manufacturing is immune. Look at what has happened recently
to once stable products like telephones, watches and even many industrial products.
Being the first in the marketplace with an improved product is clearly an obvious
competitive advantage. W hy do we stress the obvious? Unfortunately, many m anu­
facturing people have the feeling that engineering changes are often made simply to
make their life more difficult.
The impact o f engineering improvements on the market looks as if it depends only
on our market research efforts to define market needs, and the ability o f our product
engineering departments to develop the needed products. How can it be that inventory
has an impact here?

46
Product

Engineering

The purpose of engineering changes

is to improve the product!

47
 24: LOW INVENTORY, QUICKER NEW
PRODUCT INTRODUCTION
Assume that an engineering change affecting the first operation is released one month
after an order has been started in the plant. In high inventory manufacturing the first
operation is already completed. The plant manager is faced with the choice o f scrapping
or reworking this material or delaying implementation o f the engineering change until
the next production order for this product. If we elect the latter choice it will be more
than three months before we can provide the improved product to the marketplace.
How many managers have the courage to scrap such an order and how many will
choose to implement the engineering change with the next order? All of us know the
answer.
In the low inventory environment, a portion o f the order has not yet been processed
through the first operation and will not require scrap or rework if we implement the
engineering change. The superior product will be available to the market in less than
two weeks. The company with the low inventory environment has the superior product
available in the marketplace for a significant period without competition and should
be able to gain additional sales and market share. As product life cycles are continually
reduced, these effects become more and more important.
It is now quite obvious that inventory effects the competitive edge o f the product,
but how does inventory effect price in a more significant way than we already recognize
through the carrying charge channel?

48
HIGH VS. LOW INVENTORY
SYSTEMS:
ENGINEERING CHANGES
HIGH LOW
INVENTORY INVENTORY

Hill
Il I II

I I I II

■III
*
Engineering
MONTHS change one MONTHS
month after
start of order.

Improved product will Improved product will

be available only be available in less
several months after than two weeks
engineering change.

49
 25: MURPHY—VILLAIN OR EXCUSE?
Price is a well understood and sought after competitive advantage. The company
with high margins has flexibility to selectively lower prices. O r it can use its high
margins to gain a competitive edge in other ways such as increasing its sales force,
advertising or product engineering. If we can become the low cost producer, we have
a clear advantage. Unfortunately, there is often a big gap between planned and actual
margins.
M urphy’s Law (things will go wrong at the worst possible time) is well known in
all our plants. No matter how well we plan, even when we build in large measures o f
safety, we still constantly expedite and expend lavish amounts o f overtime in order to
get orders shipped on time. This problem is so widespread that it’s often referred to
as the end-of-the-month syndrome. Somehow through special efforts we ship more
than half the m onth’s production in the last few days.
W henever we find ourselves in trouble with shipping on time, whether it’s the end-
of-the-month or when an important order is due, we invariably turn to overtime,
premium freight and other expensive, unplanned actions. The end result is that we may
or may not get the orders out on time, but we certainly incur additional operating
expense and our margins shrink. Is Murphy really the one causing these delays and
thus the increased operating expense or is it the mode o f high inventory that we are
using?

50
 High Margins

Price v

Plan, take safety measures, nevertheless,

you will need overtime to finish the order on
time.
—Murphy

51
26: HIGH INVENTORY—THE REAL CAUSE OF
OVERTIME
There is no absolute measure o f a high or low inventory environment. These are
relative terms. We can only judge whether we have high or low inventory by looking
at our competitors. If we have high inventory relative to our competitors, we will have
long production lead times since work-in-process inventory and production lead time
are really the same thing. If our competitors have lower inventory, our marketing
people will probably be forced to promise deliveries in a shorter period than our normal
lead .time. Suppose marketing had to promise the order for delivery in three months,
which is less than the four month production lead time o f the high inventory company.
Manufacturing will be forced into considerable overtime and possibly other additional
costs in order to meet the delivery date.
In the low inventory environment, the production lead time is substantially shorter
than the three months demanded by the marketplace, and no overtime will be needed
even if Murphy does strike. Inventory is not generally recognized as causing overtim e,
but maybe it’s the prime reason. In the defense industry the importance o f inventory
is generally downgraded because o f government progress payments. Consequently
work-in-process inventories are very large and production lead times are quite long.
Nevertheless overtime is typically much higher than in almost any other industry.
The linkage of inventory and the competitive edge elements seems closer than we
originally recognized. There certainly is also an impact o f inventory on investment per
unit, but is there also more than meets the eye?

52
HIGH VS. LOW INVENTORY
SYSTEMS:
HIGH MARGINS
HIGH LOW
INVENTORY INVENTORY

H ill
Il I II

I I I II

■III
MONTHS MONTHS
M arketing had to prom ise the order in 3 m onths

The plant is forced into Plant production lead

considerable overtime. time is shorter than the
lead time promised by
marketing. No overtime
will be needed.

53
 27: DOES THE END-OF-THE-MONTH
SYNDROME CAUSE THE PURCHASE OF EVEN
MORE EXCESS CAPACITY?
Coping with the end-of-the-month syndrome is a major, ongoing problem for most
plants. Every month, we encounter a surge o f product at the final operations that must
be processed in the last week o f the month if we are to make our shipping goals. We
start with liberal doses o f overtime but find that it’s often insufficient to handle this
peak load. Next we find ourselves continually requesting additional equipm ent for these
last operations. There never seems to be enough machine capacity in the final operations
when we need it.
Despite the widely perceived need for additional machine capacity, studies performed
in dozens o f plants show that, in almost every case, the existing machine capacity of
the last operations is several times higher than their average load. In fact there is usually
enough capacity to handle even the very optimistic projections in the last year o f the
five-year forecast.
What causes this apparent contradiction? Is it possible that the substantial excess
capacity that already exists at the final operations and the continual pressure to add
more o f it is caused by inventory?

54
 Lower Investment
Per Unit

We cannot finish the month because

the final operations are experiencing a peak
load . . . once again!

We must buy more machines!

—A common complaint
in almost any plant.

55
 28: HIGH INVENTORY MEANS EXTRA
EQUIPMENT, SPACE AND INVESTMENT
In the high inventory environment the last operations are at peak load for an extended
period and this peak load occurs at the worst possible time. When the material finally
arrives at the last operation we have a large peak load. Yet we are forced to quickly
expedite material through these operations because o f the end-of-the-month problem.
Overtime helps but is not always sufficient. We often find that we do not have enough
machines to accommodate the peak load within the available time. We find ourselves
in an awkward situation. Even though these machines are often idle, we are forced to
invest in more machinery in order to make the monthly shipping targets.
In the low inventory environment, the load on the last operations is more uniformly
spread and the idle time is more evenly distributed, even at the end o f a month.
Consequently, we are better able to handle the expediting, if it occurs at all, without
buying additional equipment. The excess capacity required in a high inventory envi­
ronment coupled with the inherently higher inventory greatly increases the investment
required per unit of product. In fact our investments in inventories and production
facilities typically comprise more than two-thirds o f the total investment of a m anu­
facturing company. In the low inventory environment the investment in equipm ent,
facilities and inventory are much less and consequently the retum-on-investment much
higher. Even more importantly the break-even point is lower, enabling us to be much
more flexible in pricing our products.
Do these intangible impacts o f inventory also extend to the competitive edge of
responsiveness? The general impression is that we should increase, not reduce, inven­
tory in order to improve our responsivesness. Is it so?

56
HIGH VS. LOW INVENTORY
SYSTEMS:
INVESTMENT PER UNIT
HIGH LOW
INVENTORY INVENTORY

1 I I

MONTHS MONTHS

Last operation is at peak The load is more uniform

load for a long time. at the last operation, no
Pressure to ship the order additional investment is
may force plant to buy required.
more machines which will
not be utilized most of the
time.

57
 29: IS IMPROVING DUE-DATE
PERFORMANCE BEYOND A PLANT’S
CONTROL?
Almost every plant feels a need to improve its due-date performance. They also
often feel helpless since they may lack control over the factors that cause them to miss
due dates. It looks as if the major reasons for missing due dates are external to a plant.
Either the vendors are unreliable or customers are constantly changing their minds by
adding orders, canceling orders and changing due dates. One o f the most common
complaints o f a plant manager is “ just give me a reliable forecast and I’ll ship things
on time. ’’
It’s generally true that both o f these conditions exist and heavily impact the ability
o f a plant to deliver on time. But does it mean that the solution to this problem is
beyond the plant’s reach? Maybe the real solution lies in something that is totally under
the plant’s control— its level o f work-in-process inventory.

58
 Due Date
Performance

Responsiveness

We miss due dates because our

vendors are not reliable!

We miss due dates because our

customers are constantly changing
their minds!

—The two most

common excuses.

59
 30: LOW INVENTORY—THE KEY TO MORE
ACCURATE FORECASTS
In order to understand the impact o f work-in-process inventory on due dates we must
examine something that looks at first glance as totally unrelated— the validity o f our
product forecast. Almost every plant has a forecast o f demand which is quite reliable
for some period o f time into the future, then the validity o f the forecast drastically
deteriorates within a very short period o f time. W hat causes this universal phenomenon?
If all companies in an industry are providing delivery o f a product within two months,
then customers will not place orders and commit themselves to specific due dates a
year in advance. They probably will place their orders about 2 1/2 months before they
need the product. Even when they place an order for a whole year, they will feel free
to change the quantity and ship date two months in advance without risk o f jeopardizing
deliveries or placing their vendors in an impossible situation. Consequently, the plant's
forecast for this product will be quite reliable for the first two months and quite
unreliable for the period beyond three months. If we operate with high inventory
relative to our competitors, it means that our production lead time is longer than the
valid forecast horizon of the industry. The length o f the valid horizon will be dictated
by our low inventory competitors. As a result, the high inventory com pany’s production
plans are based on pure guesses and not on a reliable forecast.
It’s no wonder that due-date performance is a problem where we have high inven­
tories. When we operate in a lower inventory mode than our competitors, we enjoy
an enviable position that gives us an inherently more accurate forecast. Now when we
start production, we have firm orders or a valid forecast which is much less likely to
change. O ur due-date performance will certainly be much improved. O ur production
plans are now driven by more reliable information and we are in a much better position
to give reliable requirements to our vendors. Remember, a prime reason that our
vendors cannot deliver reliably is because we keep changing our requirements on them ,
the same way our customers are changing their requirements on us.
How about the last competitive elem ent, shorter quoted lead times? Will we again
find that inventory plays an unexpected role?

60
HIGH VS. LOW INVENTORY
SYSTEMS :
DUE DATE PERFORMANCE
HIGH INVENTORY LOW INVENTORY

H ill
Mill!

FORECAST VALIDITY
Production starts based on a Production starts based
guess. We oscillate between on good knowledge.
excess finished goods inventory Due date performance
and missed due dates. » 90%

61
 31: SHORT LEAD TIMES—A KEY TO
SURVIVAL?
Lead times are beginning to play an increasingly important role in the competitive
edge race. An excellent example is the automotive industry’s movement to just-in-tim e
suppliers. If a vendor cannot learn how to supply the automotive assemblers just-in-
time then they are not likely to remain a supplier for long. A pretty powerful reason
for learning how to cut production lead times.
We have also seen the enormous power o f shorter quoted lead times in a wide variety
of other industries. In case after case companies have had dram atically increased market
share when they had significantly shorter lead times than their competitors. In some
cases, it has been possible to actually command premium prices when quoted lead
times are substantially less than other competitors. This is a huge competitive advantage
that many W estern industries could have over foreign competitors because o f the time
required for ocean freight shipments. In these industries there should be no reason for
a foreign competitor to beat us in our own market. It seems as if shorter quoted lead
times should require more inventory, especially in work-in-process and finished goods.
The less processing remaining to complete the product, the faster our response should
be. But is it so?

62
Responsiveness

Shorter Quoted
Lead Times

My customers never complained about my

quoted lead times until competition
started an unfair war!

—The manager of a closed plant

63
 32: INVENTORY LEVELS AND PRODUCTION
LEAD TIMES ARE THE SAME THING
Production lead times and work-in-process inventory are really the same thing. One
is a mirror image o f the other. If we reduce work-in-process inventory, production lead
times are reduced proportionately. W hat is not as well recognized is that finished goods
inventory should be proportionate to work-in-process inventory.
L et’s demonstrate it by an example. If a plant has one week of work-in-process
inventory, then on the average its production lead time will be one week. Suppose that
this plant is serving a very demanding market, a market that requires immediate
deliveries. Since the plant could supply everything within one week, they should have
about I to 1 1/2 weeks o f finished goods inventory in order to meet customer demand.
Some additional protection is needed beyond the normal production lead times due to
the uncertainty of demand.
If another plant has three months o f work-in-process inventory and is operating under
the same market condition, it will be forced to hold nearly five months of finished
goods. Some companies have demonstrated that it is possible to change a make-to-
stock business into a make-to-order one by reducing production lead times sufficiently.
We stress the word should be proportionate to work-in-process inventories, and not
is proportionate, since a reduction in work-in-process does not automatically cause a
reduction in finished goods. M anagement must adjust their finished goods levels in
accordance with any new level o f work-in-process inventories to achieve these benefits.
Thus responsiveness to the market demand is directly proportional to work-in-process
inventory.
Since inventory impacts all six competitive edge elem ents, we are forced to conclude
that the carrying charge channel is not the only indirect connection o f inventory to the
bottom line measurements. There must be another indirect linkage o f inventory to our
goal.

64
HIGH VS. LOW INVENTORY
SYSTEMS:
LEAD TIMES
HIGH LOW
INVENTORY INVENTORY

MONTHS MONTHS

Quoted lead times Lead times are

are long. much shorter.

# Production lead tim es are proportional to work-in-

process inventory.
# Finished goods inventories should be proportional
to production lead tim es.

65
33: INVENTORY AND FUTURE THROUGHPUT
Our analysis o f the competitive edge elements illustrates how closely inventory is
linked to sales (throughput) . Inventory should now be associated in our minds with
future sales, with our ability to survive and thrive in tom orrow ’s markets. The more
inventory that we are holding, the less promising will be the future. The less inventory
that we hold today, the more secure our future is. We have also seen some unexpected
impacts o f inventory on operating expense, like inventory being a major source o f
overtim e, quality costs, expediting expenses and excess capacity.
These new indirect linkages have a major impact on future throughput and an
unexpected, additional impact on operating expense. We call it the Competitive Edge
Channel. Now we are facing a situation where inventory is impacting net profit twice,
and return on investment and cash flow three times. We all know the importance o f
throughput. We are equally aware o f the importance o f operating expense. It is about
time that we recognize the importance of inventory, at least to the extent that the
Japanese intuitively sense its importance.
THE COMPETITIVE EDGE
IMPACT :
OPERATIONAL MEASURES
AND THE BOTTOM LINE
RETURN ON
NET PROFIT INVESTMENT CASH FLOW

COMPETITIVE
EDGE

67
 34: WHY INVENTORY IS A SECOND-CLASS
CITIZEN
If there are so many obvious and important aspects o f inventory to our competitive
edge, there must be a very compelling and powerful reason why all companies are not
currently operating in a low inventory mode. W hat causes so many plants to be at the
other end o f the inventory spectrum? The answer lies in our short-term view o f the
relative importance o f throughput-inventory-operating expense and with the techniques
available to manage the logistical flow o f materials.
Every plant manager is painfully aware o f the short-term importance o f throughput
and operating expense. H e’s often afraid that reducing inventory will adversely affect
them. If a plant manager misses his shipping targets a couple o f months in a row by
as little as 10% the plant will probably lose money. The plant manager may be in deep
trouble. Consequently, he’s likely to keep lots o f inventory just in case it’s needed.
His other concern is that if inventory is reduced too much, some operations might be
starved for work, causing operating expenses to go up. O ur performance measurements
rivet our attention on these short-term m easures, keeping inventory high and distracting
us from their longer term importance.
Compounding our inattention to inventory has been the lack o f effective logistical
systems to drive inventory down without the risk o f losing throughput or increasing
operating expense. Consequently, we have traditionally clung to inventory as a security
blanket to protect us against the complexities and disruptions o f our plants and the
vagaries o f customer demand.
The frantic race for a competitive edge has changed all that. Currently, there is a
frantic, worldwide search for an improved logistical system. A new buzz word, syn­
chronized manufacturing, has been coined to encompass a yet undefined, better way
to manage material flow.

68
The key to reduced inventory is
synchronized manufacturing

. . . but what do we mean by

this phrase?

69
 35: WHAT IS SYNCHRONIZED
MANUFACTURING?
Synchronized manufacturing is any systematic way that attempts to move material
quickly and smoothly through the various resources of the plant in concert with market
demand. The Japanese have used the example o f a river system to characterize the
smooth, even flow they are striving to achieve. Material should flow like brooks into
streams, and streams into rivers and on and on without dams or disruptions interrupting
the flow. There are a number o f different types o f logistical systems for planning and
scheduling the procurement, production and distribution o f materials. How can we find
out if the Japanese example is sound, and which of the various available methods is
best, and in what ways it is superior?
L et’s try to approach this problem by using an analogy. We need an analogy that
we can all relate to and that will enable us to express our dilemma o f reducing work-
in-process inventory without damaging throughput and operating expense. Once such
an analogy is developed we will try to find a solution within the framework of the
analogy. W hen an acceptable solution for the analogy is developed we will then transfer
that solution into a plant environment and check its feasibility for reducing work-in-
process inventory without harming throughput and operating expense. In this way it
will be easier to examine the various logistical procedures and to compare them in a
meaningful way.
The analogy that we have chosen is a troop of soldiers on a forced march.

70
FINDING A SOLUTION

Define an Analogy in which our

problem can be easily expressed
*
Find a solution within the
framework of the analogy
*
Transfer the solution to the
plant environment
*
Check the feasibility o f the
solution

71
 36: SPREADING TROOPS MEAN HIGH
INVENTORIES
At first glance this seems to be a strange analogy, but a troop o f soldiers on a forced
march is remarkably similar to a manufacturing plant. We can view the first row o f
soldiers marching on the road as receiving raw material, virgin road, into the plant.
This material is sequentially processed, walked on, by the subsequent rows o f soldiers
(production resources) . The last row releases (ships) the finished goods, road that the
entire troop has walked on. O ur troop uses production resources to receive raw m a­
terials, process them and produce a finished product— the same as a real plant.
In our analogy, work-in-process inventory is simply the distance between the first
row o f soldiers — those that convert raw material into work-in-process, and the last
row which transfers the work-in-process into finished goods. When the troop starts its
forced march, the soldiers are tightly bunched. But after only a few miles spreading
is quite evident and continues to grow as the forced march continues. This spreading
is a natural phenomenon which is found not only in our troop analogy, but also in
activities as diverse as a funeral procession and a manufacturing plant. The spreading
is caused by the combination o f dependent events (activities that must be done se­
quentially) and statistical fluctuations. The spreading (inventory buildup) that occurs
under these conditions can be demonstrated mathematically and is described in depth
in THE GOAL.
The problem o f reducing work-in-process inventory without jeopardizing throughput
can be stated in our analogy as reducing the spreading o f the troop without reducing
its overall speed. What can we do to prevent the spreading o f the troop without slowing
down its overall movement? This is our problem.

72
A TROOP ANALOGY
MARCHING SOLDIERS

RAW GOODS
MATERIAL FINISHED

«— WORK-IN-PROCESS---------------------- »

Spreading troops mean high inventory.

Closely packed troops mean lower
inventory.

How can we prevent the troops from

spreading?

73
37: REARRANGING THE SOLDIERS REDUCES
SPREADING
L et’s borrow an idea known by every experienced troop commander. If we put the
slowest soldiers in the first row, followed by the next slowest soldiers in the second
row, and so on until we have the strongest soldiers, those who are most able to run
to close gaps, in the last row, we can reduce the spreading. In this way, whenever
spreading occurs, the strongest soldiers (production resources) are positioned so that
they can use their strength (extra capacity) to run and close the gaps (reduce work-in-
process inventory) .
The throughput — the rate at which the troop as a whole is moving — is dictated,
in any event, by the slowest soldier no matter where he is placed within the troop.
Rearranging the soldiers in this sequence will reduce spreading without impacting the
overall speed. This approach works for a troop comm ander but will it help a plant
manager?

74
A TROOP ANALOGY

Put the slowest soldiers at the front and

the strongest ones in the rear.

75
 38: A GOOD IDEA—BUT TOO EXPENSIVE
Transferring this solution into a plant environment means restructuring the plant so
that the most heavily loaded resources (slowest so ld ie rs), the ones that can barely cope
with the load, are the machines performing the first operations. Each subsequent
operation would be done by the resource having the next least amount o f excess
capacity. If we were to restructure our plant in this way, the last operations would
have the greatest amount o f excess capacity. Any waves o f inventory (spreading sol­
diers) that build up in the plant can be absorbed by the excess capacity o f the upstream
operations (the strong soldiers that can run to close up the gaps) .
It sounds like a good idea, but let's first estimate the cost, effort, time and chaos
that will result from implementing such a plan. The results are intimidating. If we look
further and consider that future changes in our product mix might alter the loads on
our production resources, necessitating another restructuring o f the plant, we are con­
vinced that this is not a feasible solution. Our good idea has turned out to be very
expensive and inflexible, so let's go back to our analogy and look for a more feasible
solution. A word o f warning — this solution looks appropriate when we are designing
a new plant. Nevertheless, even in that case, there are much better solutions.
So we are back to the drawing board in our search for a better solution to our
problem.

76
A TROOP ANALOGY

In other words, restructure your factory so that

the most loaded machines (the primary capacity
constraints) are at the first operations and have
the machines with large excess capacity
downstream.

Estimate the cost . . . and look for a more

feasible solution.

77
 39: DRUMMERS AND SCREAMING
SERGEANTS
There is another way that a troop comm ander prevents his soldiers from spreading.
He can put a drum mer in the front row to set the pace or cadence for the troop.
W henever spreading occurs, the com m ander will have his sergeants shout at the ap­
propriate soldiers to pick up the pace and close the gap. The drumbeat helps the troop
to march in unison and coupled with the sergeant’s expletives, limits the spreading.
The overall pace o f the troop is dictated by the slowest soldier. If the slowest soldier
can march to the drum beat, then the spreading o f the troops (buildup o f work-in-process
inventory) is contained without reducing the overall speed. Note that the drumbeat
constrains the stronger soldiers from marching faster, even though they have the capacity
to do so. Again, this approach works for the troop com m ander, but how can we use
a drum mer and shouting sergeants in our plants?

78
A TROOP ANALOGY
1/ /

Put a drummer at the front to set the

pace.
Have the sergeants constantly urge the
soldiers to close any gaps.

79
 40: EARMUFFS ON THE SOLDIERS
Using a drummer and sergeants in the plant initially seems strange, but isn’t it really
a common practice? The drum mer is the materials or production control manager
assisted by a computerized system. The sergeants are the expeditors. The drummer
develops plans and schedules for when material should be procured and processed
through the various production resources in order to meet customer requirements. The
drumbeat is the production schedules dictating when and what material is supposed to
be processed by each production resource. Expeditors are needed because orders are
constantly behind schedule (unplanned work-in-process inventory — spreading) and
we must push them to meet due dates — to close the gaps. Expeditors are, o f course,
not just people with this title, but frequently everyone in management.
It seems as if we are using this solution in our plants in the same way that it is used
in the analogy, but is it so? What would you think of a comm ander who put earmuffs
on the soldiers so they could not hear the drumbeat and then told each soldier to march
at his maximum speed. The strong soldiers in the front lines will be forced to walk
as fast as possible and cause spreading to occur between them and the weaker soldiers
that follow. What would you think o f such a commander?
It seems crazy but this is exactly what we are doing in our plants. Why would we
do such a contradictory thing? The answer lies in attitudes that are deeply ingrained
in our culture. What plant does not have the following motto:

80
A TROOP ANALOGY

That’s common practice now:

The sergeant is the expeditor and the drummer

is the material management system assisted
by a computer.
But can the soldiers follow the drum beat?

81
 41: “ KEEP THE WORKERS BUSY”
“ If a worker doesn’t have anything to do, let’s find him something to do. ’’
It looks as if all our work ethic is based on this premise. In our plants this motto
usually translates into giving a production w orker more material to work on so he can
produce additional products. Isn’t that the same as putting earm uffs on the soldiers and
causing each one o f them to march as fast as he can even though it increases work-in-
process inventory and doesn’t increase throughput? In our plants, don’t the use o f
efficiencies, piecework incentives and variances serve as the earm uffs on our workers.
L et’s examine this idea in more detail.

82
A TROOP ANALOGY

“I f a worker doesn’t have

anything to do,
let’s find him something to do.”

As long as this mentality exists, each soldier

will proceed according to his potential and not
according to the constraints of the troop.
Do efficiencies, incentives and variances allow
your workers to follow the drum beat?

83
 42: COULD HIGH EFFICIENCIES
EVERYWHERE BE BAD?
Suppose that worker “ X ” is the slowest soldier and that this soldier is not in the
first row. This means that “ X ” is fed material by a stronger soldier. Transferring this
concept to our plant means that a bottleneck (X) is not at the first operation but is fed
by some other non-bottleneck resource (Y) that has more capacity than the bottleneck.
O ur stronger soldier, the non-bottleneck, can produce more parts than the slowest
soldier, the bottleneck, in the same time period.
Put yourself in the shoes o f the foreman in charge o f these non-bottleneck resources.
If you are measured by “ efficiency, ” what will you encourage your workers to do?
Work all the time so that your efficiencies will be very high! What will happen to
these parts at the bottleneck? W on’t they accumulate in front o f it? Certainly a bad
move from the point o f view o f the total plant. By attempting to be efficient we created
more inventory (spreading) without gaining any additional throughput. But you
shouldn’t worry, the accumulation o f the unneeded inventory will not occur in your
deparm ent, but somewhere further along in the production process. You will be judged
as doing a very good job— attaining very high efficiency from your workers.
One the other hand, if you try to do what is good for the plant as a whole you would
restrict your workers to produce at the rate o f the downstream bottleneck, which is less
than their capacity. What will happen to your efficiency? Do you really think you will
score high against your measurements? How do you think management will judge your
performance if your department has low efficiencies? W hat would you choose if you
were the foreman?
The same thing happens to a foreman in charge of non-bottleneck resources feeding
parts to an assembly operation which also depends on parts that are in short supply.
We also have the same condition when a foreman produces parts for which the marketing
demand is less than his capacity to produce. In all three cases there is a buildup of
inventory someplace else in the plant— a spreading of the troops. W hile the forem an’s
performance looks good, he increases the overall inventory o f the plant but not its
throughput. Not exactly what we want to achieve. Is it possible that our work ethic
causes our foremen and workers to do the wrong thing?
Maybe we need to consider some cultural changes so that the workers will have an
incentive to follow the drumbeat— some way to remove the earmuffs. But even if we
were to permit the workers to follow the drum m er, are we now beating the drum in
a way that they can really follow it?

84
THE CAUSE OF
EXCESS INVENTORIES

X—The slowest soldier—a resource

that can barely cope with demand

Y—A faster soldier—a resource with

excess capacity

85
 43: CAN WORKERS MARCH TO YOUR
DRUMBEAT?
In your plant, does the drum beat according to the constraints o f the plant or is it
driven by some unrealistic assumptions? For exam ple, do you use a logistical procedure
that assumes each resource has infinite capacity— each solider can march as fast as
you wish? Not even a single slow soldier exists? If you do, then even if the soldiers
try very hard, they cannot always follow the drumbeat. O r does your drumbeat assume
that there are predetermined lead times for manufacturing products? Even though the
average production time may be three months, when necessary, we all know that we
can expedite the completion of any order in just a few days by giving it first priority
at each operation.
What is the correct lead time we should use in beating the drum— three months or
three days? Maybe the lead time through the plant depends on how we decide to
schedule production. If an order follows its normal course it takes three months. If we
give it a priority everywhere, it can be completed in a fraction o f that time. It looks
like w e’re forced to conclude that lead times cannot be accurately predeterm ined, but
are a function o f how we decide to schedule the plant.
Does your drumbeat assume that production will be done in fixed and constant batch
sizes even though we are well aware that splitting and overlapping o f batches occurs
constantly on the shop floor, especially to get shipments out at the end o f the month?
If your logistical system uses unrealistic assumptions like these, then your drum mer
has no rhythm. He beats out discordant sounds that no one can follow.
The drummer-sergeant approach looked at first like a very good solution. When we
examine more closely how it has been implemented we are forced to admit that it is
less than satisfactory. So let’s look for another solution, even if at first sight it looks
a bit radical.

86
IN YOUR PLANT . . .
Does the drum beat according to the
constraints of the plant or according to
some unrealistic assumptions . .
Like:

# Infinite capacity
# Predetermined lead times
# Fixed, constant batch sizes

87
 44: ROPE THE SOLDIERS TOGETHER LIKE
MOUNTAIN CLIMBERS
Let us tie the rows of soldiers together — as if they were mountain climbers. In this
way we can limit the spreading (inventory) to the lengths o f the ropes. This strange
idea is actually being used in plants with great success. It was first tried by Henry Ford
when he developed the assembly line. More recently, Taichi Ohno, o f Toyota, the
father of Just-in-Tim e, employed it when he developed his highly successful Kanban
scheduling system.
Ford linked together production resources by using conveyor belts, physical ropes.
Ohno used cards or logistical ropes. Both of these rope systems have proven to be
enormously successful and have had far-reaching economic implications. Ford’s system
worked well for high-volume products manufactured on dedicated equipment. Instal­
lation of Ford’s assembly lines ushered in an era o f mass production and resulted in
a huge increase in our standard o f living. O hno’s Kanban system extended Ford’s idea
to repetitively manufactured products produced on non-dedicated equipm ent. Its in­
stallation led to the emergence o f Japan as a major economic power. We are seeing
clearly the results of O hno’s approach— a substantial increase in the standard o f living
in Japan and the loss o f our dominant position in many industries.
Can we find the secret o f the ropes? Will it enable us to win the race?

88
A TROOP ANALOGY

The invention o f Henry Ford

— The assem bly line—

Dr. Ohno from Toyota

— The K anban system —

89
 45: PREDETERMINED BUFFERS MAKE IT
WORK
The key to Ford’s and O hno’s systems lies not in the conveyor belts or Kanban cards
but in the fact that the belts and the cards are mechanisms for establishing a prede­
termined inventory buffer (rope length) between each two work centers.
On Ford’s assembly line the predetermined buffer is the space on the conveyor belt
between two operations. In O hno’s approach it is the number of Kanban cards, one
for each container o f parts, that are predetermined to be used between two operations.
The buffer tells the worker at the preceding work center both when to work— and when
not to work. When the buffer is full, the preceding worker stops. When the buffer is
not full, he works. In these two similar approaches the flow o f work is synchronized
so that the inventory is quite low compared to our conventional modes of operation.
There is, however, a major drawback to this type o f rope system. Any significant
disruption at any work center will cause the overall flow to stop and throughput to be
lost. These disruptions are very expensive since if they had not occurred, additional
products could have been produced for essentially only the cost of the raw materials.
This is the reason why extensive attention is given in these rope systems to reducing
fluctuations and disruptions in the flow of material. Machines must be much more
reliable. Setup time must be reduced and made predictable. Production overloads must
be prevented, and so on. To achieve the elimination of these problems is not a trivial
task. It requires the same long process that we are fam iliar with when debugging an
assembly, process or transfer line.
Reducing fluctuations under the Kanban system is not less important, and a very
long period, sometimes more than ten years, is needed to reduce fluctuation so that
the Kanban system can safely be installed. Let us now compare our conventional
system and this rope system.

on
SYNCHRONIZED
MANUFACTURING
ASSEMBLY LINES AND
KANBANS

Predetermined inventory buffers (either limited

by space or number of cards) regulate the
rate of production for assembly lines and
Kanban systems. The instruction given to the
worker is

“Stop working when

the buffer is filled !”
The work is synchronized, inventory is low . . .
but any significant disruption will cause the
entire system to stop.

91
 46: THE WESTERN WAY—JUST-IN-CASE
The conventional W estern approach can be characterized as a just-in-case system.
The drum that dictates when raw materials are going to be released into the plant is
held by the excess capacity o f the first operation. Remem ber, when a w orker doesn’t
have any work to do, we find him more material to work on.
The result is considerably higher inventory than in the rope system, but with the
advantage that throughput is seemingly protected. Unfortunately, we protect our current
throughput at the expense o f sacrificing our competitive edge in the market— our future
throughput. The reverse is true under the Just-in-Time rope system.

92
A JUST-IN-CASE SYSTEM

Raw Material O O O O O O O O O Finished Goods

The drum is held by the excess capacity

of the gating operations

RESULT:
# Inventory is high
# Current throughput is protected
# Furture throughput is in danger

93
 47: JUST-IN-TIME OR JUST-IN-CASE—THE
ROPE OR THE AX?
In the Just-in-Time system, the drum is held by the market demand. The release of
raw materials into the plant results from a chain reaction initiated when the final
operation supplies material to the marketplace. W hen products are shipped to a cus­
tomer, the final operation replaces these goods by withdrawing and processing an
equivalent amount of material from the buffer between it and the preceding operation.
The use o f this material signals the preceding operation to replace the material that has
been taken from the buffer. This chain reaction, or pulling o f ropes, eventually causes
an equivalent amount o f raw material to be released into the plant.
This chain reaction is accomplished through some type o f signaling device or sign­
board (K an b an ). The Kanban is often a card that is placed in a standardized container
holding a specified number o f parts. When a container is taken by the next operation
for processing, the card or Kanban is returned to the preceding operation. This Kanban
is a signal for the preceding operation to produce another container o f parts to replace
the one that has been used.
In this approach, inventory is limited by the length o f the ropes— predetermined
inventory buffers— and is much lower than in the just-in-case approach. Current
throughput may be lost whenever a significant disruption occurs, but in the longer run
the lower inventory secures future throughput by increasing the competitive edge.
What should we do? Copy the Japanese and adopt the Just-in-Time approach?
Unfortunately, we don’t have the luxury o f the long and arduous period that the
installation o f such a rope system takes. If we do nothing our heads will be chopped
off by our competition. What will we choose? The rope or the ax? Maybe there is a
better alternative. Let’s go back to our analogy.

94
A JUST-IN-TIME SYSTEM

Raw Materials 0 0 0 0 0 0 0 0 0 Finished Goods

The drum is held by marketing demands

RESULT:
# Inventory is low
# Current throughput is in danger
# Future throughput is increased

95
 48: A NEW SYSTEM—DBR
Since the weakest soldier dictates the pace, if we allow the first soldier to go faster
than the weakest soldier, he will just cause the troop to spread. Why don’t we tie a
rope directly from the weakest soldier to the first row of soldiers? This is a different
approach to synchronizing our troop (manufacturing plant) , so we need a name for
it. W e’ll call it the Drum-Buffer-Rope (DBR) approach.
L et’s explore the DBR way to understand its ramifications. The soldiers following
the weakest soldier will be able to march faster than him and consequently will always
be on his heels (no spreading here) . The first row o f soldiers could also march faster
than the weakest soldier, but is constrained by the rope to march at the same speed
as the weakest soldier. The soldiers between the first row and the weakest soldier are
faster than the weakest one and thus will be on the heels o f the first row of soldiers.
The only gap or spreading will be right in front o f the weakest soldier. The size o f this
gap will be predetermined by the length o f the rope that we have chosen.
Let’s examine the advantages o f this solution. Suppose that one o f the soldiers
following the weakest soldier drops his gun. Under Ford and O hno’s rope systems,
the whole troop will soon be halted. Under the DBR way the weakest soldier’s progress
will not be affected at all. Some spreading (inventory) will occur because o f this
disruption, but since the soldiers that follow the weakest soldier are faster (have extra
capacity) , they will catch up a little bit later. The spreading will be only temporary
and there will be no slowing o f the progress o f the entire troop (throughput) . We see
that the impact of a disturbance on the DBR way is vastly different than on the Just­
in-Time approach.
It seems that the DBR way has some advantages, but let’s look further. If a soldier
preceding the weakest soldier drops his gun, as long as he picks it up before the weakest
soldier has closed the gap, there will again be no impact on the troop’s rate o f movement.
The gap (inventory) in front o f the weakest soldier serves as a buffer against disruptions
from the preceding soldiers (production resources) . By concentrating the inventory
just in front o f the weakest soldier and causing the first soldier to walk at the weakest
soldier’s rate, we are gaining the best o f both worlds. Inventory is even lower than
under Jus}-in-Time and throughput is even more protected than under just-in-case.
Maybe we have found a much better way to synchronize manufacturing. It looks
like we can protect current throughput, enhance future throughput, not endanger op­
erating expense (no more soldiers are needed) and still reduce inventory significantly.
It sounds good, but let’s check to see if the DBR way is really applicable in all
types of plants. W e’ll begin by examining a schematic o f a simple plant employing
this approach.

96
A TROOP ANALOGY

Since the weakest soldier dictates pace

# To prevent spreading, tie weakest

soldier to the front row.
# To protect overall pace, provide some
slack in the rope.

97
 49: DRUMS, BUFFERS AND ROPES
In any plant there are only a few capacity constraint resources (CCR’s) — weakest
soldiers. The DBR way recognizes that such a constraint will dictate the rate of
production o f the entire plant. So let’s treat the major capacity constraint resource as
the drummer. Its production rate will serve as the drumbeat for the entire plant. We
will also need to establish an inventory buffer in front o f each CCR. This buffer will
contain only the inventory needed to keep the CCR busy during the next predetermined
time interval (from now on we will refer to such a buffer as the time buffer) . Con­
sequently, this time buffer will protect the throughput o f the plant against any disruption
that can be overcome within the predetermined time interval.
In order to ensure that inventory will not grow beyond the level dictated by the time
buffer, we must limit the rate at which raw material is released into the plant. A rope
should be tied from the CCR to the gating (first) operation. In other words, the rate
at which the gating operation will be allowed to release material into production will
be dictated by the rate at which the CCR is producing.
The concept seems sound, so let’s devise a procedure to implement the DBR logistical
approach, the drum -buffer-rope system, in a plant. A good logistical system should
have the means (plans and schedules) o f controlling the flow o f material into, through
and out o f our plants no matter how complicated they may be. It is such a procedure
that we need to develop.

98
SYNCHRONIZED
MANUFACTURING
THE DRUM-BUFFER-ROPE
WAY

Raw Materials Finished Goods

oooo«ooo
A rope tying the Major Capacity Constraint
gating operation to Time Buffer
the buffer

99
 50: DEVISING A DRUM-BUFFER-ROPE
SYSTEM
W e’ll begin by examining a part going through several operations with only one of
them being a CCR. This part will eventually be assembled with other parts into a
finished product for shipment to several different customers.
Since the two major constraints on the plant are the market demands (the amount
o f product we can sell) and the capacity o f the CCR, it will make sense to base our
schedule (logistical flow) on those two constraints. Thus the first step will be to
determine the schedule o f the CCR by taking into account only its limited capacity and
the market demands that it is trying to satisfy. Once the C C R ’s schedule is established,
we need to determine how to schedule all the non-constraining resources. Using the
schedule o f the CCR, the schedule o f the succeeding operations can be derived easily.
Once a part is completed at a CCR it can be scheduled to start at the next operation.
Each subsequent operation, including assembly, is simply started when the previous
operation finishes. In this fashion we can generate schedules for all succeeding op­
erations, including assembly.
The challenge now is to schedule the preceding operations and to protect the CCR
from disturbances that might occur at the preceding resources. As we said before, we
would like to limit the buffer to a specific time interval. Let’s suppose that most of
the disruptions at the preceding operations can be overcome within two days. If so,
three day’s protection in our time buffer looks like it will be sufficient. Now we simply
have to schedule backwards in time from the CCR. We will plan for the operation
immediately preceding the CCR to complete the needed parts three days before they
are scheduled to be run on the CCR. Each o f the other preceding operations will be
back-scheduled in a similar manner so that the parts are available just-in-tim e for the
following operation.
In this way we can generate a schedule and a time buffer that will satisfy all the
requirements laid out in the schematic approach. Any disturbances at the preceding
operations that can be overcome within the time buffer will not effect the throughput
o f the plant. Looks good— throughput is protected, inventory reduced and operating
expense is not increased. How do we now schedule other parts fed to the same assembly?

ion
SYNCHRONIZED
MANUFACTURING
THE DRUM-BUFFER-ROPE
WAY
ORDERS

ASSEMBLY

C.C.R. OPERATION

The Capacity Constraint Resource

should dictate the schedule based
on market demand and its own
potential.
The schedule for succeeding
operations (including assembly)
should be derived accordingly.
# The schedule of preceding
operations should support the time
buffer and thus be derived
backwards in time from the C.C.R.
schedule.

101
 51: ASSURING HIGH DUE-DATE
PERFORMANCE
The procedure laid down so far will protect the throughput o f the plant, but meeting
customer due dates is also important and needs to be protected as well. In the DBR
way, the assembly schedule is dictated by the availability o f the scarce parts coming
from the CCR. The availability o f these scarce parts controls when we can assemble
and ship products. Consequently, we should try to prevent the shortage o f any other
part from disrupting our assembly schedule.
In order to assure that the other parts are available when needed, let's once again
build a time buffer, this time in front o f the assembly operation requiring a part from
a CCR. The purpose o f this time buffer is to protect the assembly schedule against
disruptions that might occur in procurement and manufacture o f the parts that do not
go through a CCR.
Following this approach, the schedule o f these parts should be derived backwards
in time starting from the time buffer in front o f assembly. Assume again that we have
established a three-day buffer in front o f this particular assembly operation. The parts
coming from the operation immediately preceding assembly should be scheduled to
be completed three days before they are needed at assembly. Using this back-scheduling
approach we can establish when each o f the other preceding operations should be
started and completed and when material should be received at our plants.
Now any disruption that occurs at any o f the vendors or work centers feeding the
other parts to assembly will not affect the due-date performance o f the plant if the
disruptions can be overcome within the established time buffer interval. Notice that
time buffers are not required before every assembly operation. They are required only
before assembly operations that are fed by both CCR and non-CCR parts and in front
o f the CCR itself. In this way, every part will cross, in its journey from raw materials
to finished goods, no more than one buffer.
It seems as if the DBR logistical approach applies to both flow plants and to more
complicated ones producing assemblies. A diagram illustrating a drum -buffer-rope
system for a plant making several assembled products looks like a big spaghetti diagram.

102
SYNCHRONIZED
MANUFACTURING
THE DRUM-BUFFER-ROPE
WAY

• AH other operation schedules must

support the assembly schedule.
• To avoid harm due to disruptions,
they should support a time buffer
in front of any assembly that uses
a C.C.R. part.

• Inventory is low, but nevertheless

any disruption that can be overcome
within the buffer time will not affect
the throughput of the plant.

103
 52: UNIVERSAL APPLICATION
In any plant, no matter how large or complex, there are only a limited number of
C C R ’s. Every CCR can be protected by a time buffer and so can the assemblies fed
by them. Ropes can be tied from each buffer to the gating operations and any fork
points. It looks as if there is no limit to the application of a drum-buffer-rope system.
The concept o f the DBR logistical system is quite clear, but the complexity o f this
diagram illustrates why we will need the aid o f a computerized system. Even though
the calculations are quite straightforward, to perform them manually in almost every
plant is very time consuming and requires heavenly patience.
In implementing such a procedure, the first question that jum ps into our mind is
how we can quickly identify which o f the production resources are CCR ’s.

104
SYNCHRONIZED
MANUFACTURING
THE DRUM-BUFFER-ROPE
WAY
ORDERS

ASSEMBLY

O MANUFACTURING • OPERATIONS
OPERATIONS DONE BY C.C.R.

. . . r TIME BUFFERS .... ROPES

IN ANY PLANT THERE ARE VERY FEW

CAPACITY CONSTRAINT RESOURCES.

105
 53: LOCATING THE CONSTRAINTS
At first glance, locating the C C R ’s seems to be a formidable and possibly unending
task. True, in some plants there is a good understanding o f which operations are CCR ’s.
In other types of plants the C C R ’s are referred to as floating bottlenecks that wander
all over the plant. In still other types o f plants, the locations are not at all clear and
it appears that a tremendous amount o f time and research would be needed to locate
them.
While our task initially appears to be a mammoth one, there is a solution. The key
is in recognizing that a constraint o f the entire plant (e. g. a CCR) must manifest itself
in every aspect o f the business. Building on this insight, we can devise a straightforward
approach that enables us to take surprisingly little time to zoom from the various
impacts o f a constraint directly to its location.
This method, even though it is well-defined, thoroughly tested and routinely taught
to others is beyond the scope of this book. So let’s assume at this stage that such a
method exists and that we have already applied it to determine the location o f the
C C R ’s. Once the C C R ’s have been identified, the next question is exactly how to
schedule them in accordance with their limited capacity and the market demands they
need to satisfy.

106
SYNCHRONIZED
MANUFACTURING
LOCATING THE
CONSTRAINTS

First step toward synchronized manufacturing

is to identify the constraints.

A capacity constraint manifests itself in all of

the major business issues.

An analysis of the major business issues can

be used to identify the capacity constraint
resources (CCR’s ).

107
 54: HOW TO BEAT THE DRUM
A CCR limits the throughput o f the plant and controls due-date performance. We
have to ensure, on the one hand, that the CCR is not scheduled to produce more than
it’s capacity and on the other hand not to waste any o f its capacity by allowing any
slack in its schedule. Finally, we have to sequence production at the CCR in a manner
that will result in good due-date performance.
We can accomplish these goals by employing a method that is used by almost every
foreman. First, simply schedule forward in time from the present. Decide what product
to schedule first, how many are needed and how long it will take to produce them.
Then repeat the procedure. When the available capacity for the first day is used up,
begin scheduling the second day, and so on. The only remaining problem is how to
choose the sequence in which the various products are to be done by the CCR. A good
first rough cut sequence is given by the customer due dates o f the required products.
We probably would like to work first on a product needed by a customer two days
from now before working on a product needed next week.
Sequencing production at the CCR based on the custom er due date is a sound
approach, but there are four cases that might cause us to modify these sequencing
decisions.

108
SYNCHRONIZED
MANUFACTURING
BEATING THE DRUM

Ensure maximum throughput through

forward scheduling of the C .C .R .’s.

Due dates give us the first, rough sequence,

but the sequence must be modified under
any one of four conditions . . .

109
 55: FOUR CONDITIONS THAT COMPLICATE
SCHEDULING
The first case occurs when lead time from the CCR operation to the completion of
the product is greatly different for different products. We might have a Product A that,
once it has been completed by the CCR, requires only one day o f additional work
before it can be shipped. Product B may require an entire week at the operations
following the CCR before it can be shipped. When these circumstances exist, it may
make sense to modify the CCR customer due-date sequence at the CCR so that we
work on Product B which is due for shipment next week before working on Product
A even though it is due to be shipped this week.
A second condition that may cause us to deviate from the custom er due-date sequence
arises whenever one CCR is feeding another CCR. In such a case, by obeying the
market due-date sequence at the first CCR we may starve the second CCR. We don’t
have to lose time on all C C R ’s in order to lose throughput o f the entire plant. To lose
time on just one o f them is enough. Remember the value o f this lost time. If it had
not been lost, we could have shipped additional products for essentially the raw material
costs.
A third common situation occurs when the process on a CCR also involves
setup— effort and time required for the resource to be changed from producing one
product to another. In such an event, we will sometimes prefer to make a single
production run to satisfy the market demands o f a particular product for several days,
saving several setups, rather than follow the exact custom er due-date sequence. In this
way we can use more o f the scarce CCR capacity for actual production and less for
setting up.
There is a fourth common situation that is less recognized but no less important. It
occurs whenever a CCR is producing more than one part for the same product. In this
case, the customer due date w ill'ftot guide us at all in choosing the sequence, since
all the parts have the same due date. Nevertheless, the resulting performance o f the
plant can be greatly influenced by the sequence we choose.
The task of choosing a good sequence in each o f these four cases is more complicated
than setting the sequence based solely on customer due date. However, good rules can
be established and incorporated into a computerized system . But we must stress that
the real importance lies more in the overall application o f the drum-buffer-rope meth­
od than the precise way that the drum is beaten.
The technical details of constructing such as a system can be, and have been,
overcome. The real difficulties that can prevent a company from fully and quickly
capitalizing on the drum-buffer-rope approach do not lie in the technical details. They
lie in the fact that the drum -buffer-rope method is in direct conflict with some deeply
rooted behavior patterns.

no
SYNCHRONIZED
MANUFACTURING
BEATING THE DRUM

Complicating conditions:

# Different lead times from capacity constraint

resources to due dates

# One capacity constraint resource feeding another one

# Set up on a capacity constraint resource

# A capacity constraint resource feeding more than one

part to the same product

111
 56: DBR BUFFERS CLASH WITH OUR
CULTURE
We have agreed that it seems logical to concentrate all the inventory protection just
before the crucial operations and not to use inventory to protect every place where a
disruption can occur. While this seems logical, it is contrary to the behavior o f almost
every foreman. Foremen are accustomed to buffering themselves with floats o f inven­
tory so that they can respond to any urgent demand from succeeding operations, which
they know all too well will occur at the worst possible time.
A very persuasive education must be given to the foreman before he will abandon
this long-held pack rat mentality. Remember we are asking him to give up his visible
protection, one that he controls, in return for a promise that somewhere, maybe not
even in his department, inventory is kept to protect the entire plant. We are not dealing
here with a change in the forem an’s personal culture, but a culture controlled by how
management measures a foreman’s performance.
O ur buffer only-at-critical operations concept clashes not only with a forem an’s
culture, it clashes even more strongly with the culture o f upper management. We have
just agreed that inventory o f the right parts, in the right quantities, at the right time,
in front o f the right operations, gives very good protection. We have even described
a procedure to implement this concept. We have also agreed that because o f the frantic
race for a competitive edge that work-in-process inventory anywhere else is destructive.
These two ideas, while logically compelling, are very much at odds with current
management thinking about inventory. Managements need to reanalyze and reassess
their reasons for holding inventory. The perceived reasons may be at odds with the
financial ramification of the competitive edge analysis outlined previously. A deep
soul-searching is needed to overcome the accepted practices o f many generations.
The buffer concept flies in the face o f our ingrained culture, but its implications
pale in light of the impact o f the rope concept.

112
SYNCHRONIZED
MANUFACTURING
LOCATING THE
TIME BUFFERS

0 Concentrate protection not at the origin of a

disturbance, but before critical operations

0 Inventory of the right parts in the right

quantities at the right times in front of the
right operations gives high protection

# Inventory anywhere else is destructive

113
 57: DBR ROPES REQUIRE A CHANGE IN
MANAGEMENTS BEHAVIOR
We can readily accept the logic that requires us to release and process materials
according to the schedule determined by the plant constraints (the rope c o n c e p t). Once
we do, we have to face the fact that this conclusion means that under no circumstances
should we release materials just to supply work to workers. This is probably the most
difficult behavior pattern that has to be overcome.
The Japanese have an advantage, a lead in the competitive edge race, because they
went through this cultural shock a decade ago. Imagine an expensive machine run by
a highly paid worker with some expensive, half-processed parts sitting in front o f it.
These parts are going to be needed to assemble customer orders in just four hours.
Nevertheless, the worker is not running the machine. He is standing idle.
What will be your response to such a situation? Once you have recovered from your
heart attack you probably will deal very severely with the worker and his foreman.
The Japanese response in Just-In-Time plants is quite different. Under a Kanban system
(the JIT scheduling sy s te m ), as long as this worker does not have a Kanban card, this
is precisely what he is supposed to do— not produce parts. This difference in behavior
is not because o f a difference in the culture of countries or workers but is due to a
drastically different culture o f management.
The Japanese have recognized and proven the tremendous benefits o f such a man­
agement cultural change. We must now make the same behavioral change or withdraw
from the race. It is m anagem ent's task to create the cultural change needed for the
acceptance o f these concepts. The procedure for creating the necessary supporting
schedules and simulating (what ifs) the flow o f material and* use o f resources can be
readily accomplished through computer software.
The drum-buffer-rope concepts are straightforward, readily understood and to a
degree can be capitalized on without the use o f a software system. The need for the
assistance o f software increases with the amount o f data, changes (e. g. forecast
variations) and what ifs that must be dealt with. It may become a necessity as the
number of C C R ’s increases and the four complicating conditions become more prev­
alent. It is probably essential if a company wishes to establish a focused process o f
ongoing improvement.

114
SYNCHRONIZED
MANUFACTURING
ROPES

9 Release and process material according to

the schedule determined by the plant
constraints

0 Do not release material in order to supply

work to workers

115
 58: DBR AND A PROCESS OF ONGOING,
FOCUSED IMPROVEMENT
Installing a drum-buffer-rope system can get a company back into the competitive
edge race. In fact the results o f such installations and the relatively short time required
to achieve outstanding benefits are truly impressive. However, a drum -buffer-rope
system will not enable a company to stay in or lead the race for long.
We must not look for just an improvement in our performance, no matter how
significant. We must find a way to establish an ongoing, never ending process of
improvement. We know that we need to improve in our plants. Unfortunately there
are an almost unlimited number o f improvements possible. We know that we can 't do
everything at once, so where do we start? W hat specific improvements will move us
closest to our goal? It would be even better if we had a procedure that could be used
routinely to determine which improvements are the most important at any point in time.
There is such a process o f ongoing, focused improvement.
We have seen the importance o f the capacity constraints to throughput and inventory
and why they need to be buffered against the many daily disruptions that occur in a
plant. A careful observation o f these buffers can tell us a great deal about the inevitable
fluctuations in our plant and its marketplace. Understanding how to properly manage
the inventory buffers can both improve our immediate competitive edge position (since
most of our inventory is now in the buffers) and serve to pinpoint the improvements
most needed to further enhance it.
While this process o f ongoing im provement can be used to focus and synchronize
almost all management efforts in a manufacturing com pany, we will illustrate here
only an example o f how it can be used to focus process improvements.
The following illustrations demonstrate how to manage the buffers and to use them
as a crystal ball to locate and quantify the importance o f the disruptions in a plant.
Correcting those highlighted disruptions, continual usage o f the drum-buffer-rope ap­
proach to synchronize flow and manage the buffers will enable p s to establish an
ongoing, focused process o f improvement— a productivity flywheel.
In order to better understand how, let's first look in more detail at a time buffer.

116
SYNCHRONIZED
MANUFACTURING
THE DRUM-BUFFER-ROPE
WAY

In the midst of a competitive edge race

we should not look for an improvement,
we should look to implement a process
of

on-going improvement.

117
 59: UNDERSTANDING TIME BUFFERS
Suppose that the schedule has been set for a CCR for an entire week. This schedule
calls for the processing o f various parts in various quantities throughout the days of
the week. The same part may appear more than once on this dispatch list since the
sequence is set by customer due dates and modified whenever any o f the four com ­
plicating conditions exist.
If we stick to our example o f choosing a time buffer o f three days, on Monday
morning we expect to find in front o f the CCR all the parts that the CCR is scheduled
to do on M onday, Tuesday and W ednesday. We do not want to find in front o f the
CCR any other parts. Accumulating more parts there will not add significantly to our
protection and it will reduce our competitive edge in the market.
We have chosen to depict the time buffer as a rectangle. The vertical axis represents
the number o f hours that a particular part will require o f the CCR. The horizontal axis
measures when (e. g. what day) these parts are scheduled to be processed by the CCR.
At any point in time, we have a fixed predetermined time buffer, but the content
o f the time buffer, as we shall see, is continually changing.

118
SYNCHRONIZED
MANUFACTURING
THE TIME BUFFERS
A schedule of CCR
DAY PART QU. HOURS
MON. A 25 5
B 5 3

TUES. B 5 3
C 5 5

WED. C 2 2
D 2 6
Choosing a buffer of three days.
THURS.. D 1 3 The planned buffer content on
A 25 5 Monday morning.
FRI. C 2 2
B 10 6
8'
B c D
5pcs 5pcs 2pcs
3hrs 5hrs 6hrs
HOURS OF
WORK 4" A
AVAILABLE
TO THE 25 pcs B
CCR 5hrs 5pcs C 2pcs
3hrs 2hrs

1 2 3 DAYS

No other part should be in

front of the CCR.

119
 60: BUFFER CONTENT—CONSTANTLY
CHANGING
The content of our buffer on Tuesday morning should be different. We now expect
to find only the parts that the schedule o f the CCR calls for on Tuesday, W ednesday
and Thursday. The parts scheduled for Monday should have already been completed
and the parts scheduled for processing by the CCR on Friday should not have arrived
yet. This concept of revolving inventory in the buffer is vastly different from the usual
understanding of safety stock as a constant inventory level for each part.
We have chosen to portray the buffer content in the form of a rectangle. By presenting
it in this way, we can see the quantity o f parts needed, the sequence in which they are
going to be consumed and the amount of hours o f work o f the CCR that they are
protecting. This approach enables us to analyze the actual and planned content o f the
buffer and to determine actions that will improve both our immediate and longer term
competitive edge.

120
SYNCHRONIZED
MANUFACTURING
THE TIME BUFFERS
A schedule of CCR

DAY PART QU. HOURS

MON. A 25 5
B 5 3

TUES. B 5 3
C 5 5

WED. C 2 2
D 2 6 The planned buffer content on
THURS. D 1 3
Tuesday morning.
A 25 5
FRI. C 2 2
B IO 6

HOURS OF
WORK
AVAILABLE
TO THE
CCR

The buffer content is changing from one day

to the other in accordance with the CCR
schedule.

121
 61: ACTUAL BUFFERS SHOULD DIFFER
FROM PLANNED BUFFERS
How can an analysis o f the time buffer point the way to actions that immediately
improve our competitive edge? W e know that one purpose o f the buffers is to protect
the throughput and due-date performance o f the plant against the impact of disruptions.
If disruptions occur, we should expect to find that the actual content o f the buffer is
smaller than we planned. If the buffer is always full, it is a sure sign that there are no
disruptions significant enough to affect the planned material flow. Consequently, a
buffer is not needed and this inventory can be eliminated without damage to throughput
or operating expense. In fact, by eliminating the buffer we will reduce operating
expense.
If the actual buffer in front o f a critical operation should not be the same as the
planned buffer, what should it look like? The desired planned and actual buffer pattern
is illustrated in the facing example. The material planned to be in the first third o f the
time buffer, the material that is consumed first by the CCR, should always be present.
On the other hand, we should expect to find that most of the material planned for the
last third o f the buffer is missing. The actual versus planned contents o f the middle
third o f the buffer should lie somewhere in between these two extremes. This buffer
profile should protect our critical operations from all but the most extreme fluctuations.
L et’s see what actions we can take to improve our competitive edge immediately
if the actual content o f the buffer deviates from this pattern.

122
SYNCHRONIZED
MANUFACTURING
TIME BUFFERS

The time buffer contains most of the

inventory and should protect the plant
against disruptions

----- PLANNED
...... ACTUAL
HOURS OF
WORK DONE
BY C.C.R.

¥
TIME

The actual buffer must be smaller than

planned if disruptions exist.
Otherwise, there is no need for a buffer
at all

123
 62: MANAGING THE TIME BUFFERS
If the pattern of the actual buffer stretches beyond the planned buffer, as in Case
1, then it is a clear indication that material is being produced earlier than called for
at preceding operations. This pattern suggests that material is being released prematurely
to the first operation. Management probably has not totally implemented the cultural
change required for a drum-buffer-rope system. More education and discipline is prob­
ably needed at the gating operation.
If the buffer is almost full as in Case 2, we have a clear indication that the planned
buffer is too large. We are paying too high a premium for insurance. We should cut
the size o f the planned buffer to the point where only the first third o f the buffer is
totally filled. In Case 3, the totally filled portion is less than the first third o f the buffer,
indicating that the buffer is too small and we run the risk o f starving the CCR and
losing throughput. The planned buffer needs to be im mediately increased until the first
third is completely full.
We can see why it is important to have the appropriate amount o f emptiness in the
buffer and where it should be located, but we should also be taking action to eliminate
vacancies or holes in the buffer. If we can prevent these holes from occurring then we
can further reduce the buffer size and increase our competitive edge. How can we get
an indication where, with the minimum amount o f effort, we can have the maximum
impact on reducing the buffer?

124
SYNCHRONIZED
MANUFACTURING
BUFFER MANAGEMENT

CASE 1

CASE 2

CASE 3

125
 63: HOLES IN THE BUFFERS
A comparison o f the planned and actual buffers will reveal the parts that are missing
from the buffer that should be there. These parts, or holes, in the buffer are caused
by disruptions in the flow o f material at preceding operations or from our vendors.
At this point, we do not know where this material is or why it is late to the buffer.
We do know that it must still be at one o f the preceding operations (or a v e n d o r). O ur
illustration shows that a certain quantity o f part A that was planned to be in the buffer
has not yet arrived, creating a hole in the buffer. These missing parts will require
4tY ” hours o f the OCR’s capacity when they do arrive and are processed. We also
know from our buffer that the schedule o f the OCR calls for these parts to be processed
44W ” hours from now.
We know the size o f the hole in the buffer (its impact on the OCR) and how much
time we still have to fill this hole without damaging our schedule. This information
will help us to quantify the importance o f the disruption.
What we don’t yet know is where the parts are, why they are late and what needs
to be done to rectify this problem.

126
SYNCHRONIZED
MANUFACTURING
FOCUSING IMPROVEMENTS

Discrepancy between planned and actual

buffers reveals the disruptions to material flow

Hours of
CCR
\ Y
HOU is
*4

«— w ----- ►
A hole representing Y hours of Part A
inventory scheduled to be worked on in
Whours are missing.

This information can be used to quantify the

disruption.

127
 64: CALCULATING A DISRUPTION FACTOR
We can determine the location o f the missing parts by checking our inventory control
system or by simply going and looking. Once we have determined the location o f these
parts, we have a very good indication o f which work center or vendor caused the
disruption in the flow. The disruptive source is most likely the work center or vendor
that the part is currently sitting in front of.
Now that we know the location o f the material, we can devise a procedure to quantify
the importance o f this disruption relative to other disruptions causing holes in this and
other buffers. Only three parameters are needed and we already know two o f them.
The first is “ Y, ” the number of hours that the CCR should spend processing these
parts. This param eter reflects the magnitude o f the damage that will be caused if the
material will not arrive at the buffer on time. The second is the protection time “ W ”
that is still left until the CCR will be impacted by the absence of this material. The
third one is “ P, ” the amount o f processing time required to complete the parts so
they can be processed by the CCR.
Using this information we can calculate a single number, or disruption factor for
each hole in the buffer and assign it to the disruptive work center. The larger the
disruption factor, the more important it is to eliminate the source o f the disruption.
Bear in mind that in the case that the material is not in the plant, this disruption factor
corresponds to a particular vendor.
We now know the relative importance o f all the disruptions in our plant. How do
we take advantage of this information to increase our competitive edge?

128
SYNCHRONIZED
MANUFACTURING
FOCUSING IMPROVEMENTS

Trace location of inventory Buffer W *

missing from buffer

P The processing time

This will help identify the still left to complete
source of disruption the parts

Disruptive source
P, W and Y can used to
quantify the magnitude
A—Inventory location
of the disruption

129
 65: THE PARETO PRINCIPLE
We can repeat these calculations for every hole in every buffer in our plant. If we
sum these disruption factors for each source we can arrive at a “ disruption factor” for
each resource and vendor. The size of the disruption factor tells us the importance of
the source in disrupting the flow of materials. We now know not just the relative
importance of each disruption, but also the relative importance of each disruption
source.
These work center/vendor disruption factors become our priority list for pinpointing
where we should concentrate our productivity improvement efforts. O f course, we
should deal first with the work center/vendor that has the largest disruption factor. Even
if it is difficult to analyze and correct those disruptions, we certainly should not be
distracted and correct an easy problem at a work center further down the list. The only
result of correcting such a disruption will be the satisfaction that we are correcting
something, but it will not have any meaningful bottom line impact.
Our improvement efforts should be driven by the Pareto principle. Pareto claimed
that there were always a few important things and many trivial ones. The facing
illustration portrays Pareto’s idea by contrasting the benefits resulting from an im­
provement with the cost o f making that im provement. Clearly, the most desired im­
provements are where the benefits are large relative to the cost. Constantly striving to
eliminate disruptions that cause holes in the buffers is a process o f ongoing improve­
ment. Gearing our efforts to the sources with the largest disruption factors is part of
this process o f ongoing, focused improvement.
Once we locate and quantify where we should concentrate, we have a whole host
o f good techniques to analyze and correct the problems.

130
SYNCHRONIZED
MANUFACTURING
FOCUSING IMPROVEMENTS

Repeating the same process for every

hole in every buffer and summing the
magnitude of the disruption for each
work center will give us a disruption
factor for each work center in the plant

These disruption factors are our priority

list for focusing our productivity
improvement efforts.

131
66: MAKING SENSE OF THE ALPHABET SOUP
Our disruption factors tell us where to concentrate and how important it is to make
the improvement. They do not tell us what caused the disruption. We must analyze
the source.to determine the major cause of the largest disruption factors.
A work center might be disrupting the flow because o f frequent machine breakdowns.
Here is where our preventative maintenance efforts should be focused. In the case o f
an old, unreliable machine, maybe a new one should be acquired or the largest cause
of the disruption might instead be a quality problem. Dr. Deming, Dr. Juran and others
have provided us with a whole host o f very powerful techniques on how to isolate and
resolve quality problems.
It might also be that the disruption is caused by a long and unreliable setup. Here
is where we should use the detailed setup reduction techniques developed by the
Japanese. O r it could be that the disruption was caused by a foreman trying to make
his department look efficient. He might be running bigger batches than required, in
order to save setups, but as a result causing a disruption in the flow o f needed material.
We might deal with this problem by using the “ old Missouri mule technique. ”
These illustrations are examples o f how this process o f ongoing improvement can
focus our alphabet soup o f available improvement techniques into a powerful, coherent
force. Each o f the improvement techniques can be either extremely beneficial (if it has
a global impact) or a waste o f money (where it has only a local im p a c t). Since inventory
is closely linked to the six competitive edge elem ents, we can use the time buffers to
pinpoint the most critical areas for improvement. W e should then apply the appropriate
technique, and continually repeat this process at the next most important point. We
should not apply any or all o f these techniques everywhere.
There are a number o f other causes o f disruptions, and equally effective techniques
available to eliminate them. Correcting the problem o f the work center (or the vendor)
appearing highest on the list will jjave the greatest impact on filling the most important
holes in the buffer, enabling us to reduce the buffer size and repeat the process. As
the buffers are reduced, the impact is sure to come.

132
SYNCHRONIZED
MANUFACTURING
FOCUSING IMPROVEMENTS

The work center with the highest disruption

factor must be analyzed for causes:

# Maintenance
# Quality
# Long unreliable setups
# Other

Once improvements occur, the major holes will

disappear and the time buffers can be reduced.

133
 67: REDUCING DISRUPTIONS TO GAIN A
COMPETITIVE EDGE
Focused application o f the right productivity improvement technique reduces dis­
ruptions and eliminates the most important holes in our buffers. As the buffers are
decreased, since they contain the majority of the work-in-process inventory, the com ­
petitive edge o f the plant is increased. Lead times, operating expense and inventory
investment will decrease while quality, due-date performance and the speed o f intro­
ducing improved products will increase.
The market will respond with increasing demand that will lead to a throughput
increase. This additional throughput should be very profitable since a commensurate
increase in operating expense and inventory is not required. Net profit, return on
investment and cash flow will increase simultaneously. We will be moving in the
direction o f our goal.
However, the elimination o f the most important sources of disruption and the in­
creased volume will change our plant and how and where we should focus our efforts.

134
SYNCHRONIZED
MANUFACTURING
FOCUSING IMPROVEMENTS

Buffer Decreases
T
WIP Inventory Decreases
T
Competitive Edge Increases
T
Throughput Increases
▼

r Net ▲ Return On a Cash a ^

Profit r Investment T Flow ^
s________

135
 68: DEALING WITH BOTTLENECKS
The throughput increase will reduce the excess capacity in the plant, causing the
time available to recover from disruptions to decrease. We have been protecting our
plant against disruptions with both our inventory buffers and the excess capacity o f our
stronger soldiers. The stronger soldiers now will have less capacity to catch up quickly
if they drop their guns. Now disruptions might cause the slowest soldier (CCR) to stop
from time to time (a loss o f throughput) . We now need to increase the buffers we
have worked so diligently to reduce. Plant personnel need to constantly work on
reducing the buffers by eliminating the most important sources of disruption no matter
how much the buffers are flexed to account for changing market demands and plant
conditions.
Throughput might increase to the level that real bottlenecks that limit throughput
appear in the plant. There is not yet a need to immediately rush out and buy more
equipment to increase the capacity o f these bottlenecks. There are a whole host of
quicker, less expensive steps that should be taken first. For example, we should be
certain that a bottleneck is always manned, even during lunch hours, other rest breaks
and shift changes. Care needs to be exercised so that a bottleneck does not work on
parts that are already defective (even if we have to put an inspector in front o f the
bottleneck) because wasted bottleneck time is lost throughput for the plant. Operations
following a bottleneck need to be instructed to process bottleneck parts very carefully
because every damaged part represents another lost shipment. We can squeeze a sub­
stantial amount of additional capacity from our bottlenecks by such inexpensive and
effective methods. Only when these techniques have been exhausted should new ca­
pacity be purchased.
The effort to reduce buffers and increase the bottleneck capacity on an ongoing basis
is a very profitable one. By shifting our focus from just reducing the most important
disruptions to also increasing throughput at the bottlenecks, we have raised performance
to a new level. We are beginning fo establish a productivity flywheel.

136
SYNCHRONIZED
MANUFACTURING
FOCUSING IMPROVEMENTS

Throughput Increases
e' N
Less excess capacity Bottlenecks that limit
to cope with disruptions increased production

l V
Buffer Increase Inexpensive alternatives
to buying more machines

The need to continue Capacity increase

to reduce the Buffers

C Net a
Profit I
Return On <
Investment
Cash
Flow t)

137
 69: THE PRODUCTIVITY FLYWHEEL
The first step in establishing such a productivity flywheel is to implement synchro­
nized manufacturing using the drum -rope-buffer approach. Then we need to manage
the inventory buffers and to focus our process improvement efforts. Finally, Just-in-
Time techniques, new technology and good management practices should be brought
to bear where they will have the greatest impact. The result will be a continuous
increase in net profit, return on investment and cash flow.
Establishing an ongoing, focused im provement process requires that we understand
the location of the real constraints o f our plants. Once we do, we should zoom in on
these constraints applying all our efforts to break them. When a constraint has been
removed, we have a new plant and our efforts should now be spent in other areas. We
need to find where the new constraints are and attack them with the same fierceness.
Remember, even if the constraints are external to the plant (e. g. we have plenty o f
capacity but not enough market demand) , it is still within our power to affect them.
Driving work-in-process inventory down will increase our competitive edge and cause
our market demand to increase.
This continual effort to find the current constraints, break them, find the next con­
straints, break them and so on is an extremely powerful process o f ongoing, focused
improvement. It is a way for W estern industry not only to get back into the race but
also to leapfrog over our competition.
We should use this process o f ongoing improvement to drive the productivity flywheel
at an ever-increasing speed.
BUT REM EM BER. . .

138
THE
COMPETITIVE EDGE RACE
PRODUCTIVITY FLYWHEEL

SYNCHRONIZED
MANUFACTURING

NPTN
f ROD BUFFER
MANAGEMENT

LOCAL
PROCESS
IMPROVEMENTS

JIT
/
TECHNOLOGY
MANAGEMENT

139
 70: THERE IS NO FINISH LINE
The race for a competitive edge is akin to m an’s progress— it should be ongoing and
without end. We can always do better. When we gain and appply a better understanding
o f how our manufacturing world works, many benefit. Progress in manufacturing and
a rising standard o f living have marched hand in hand since the start o f the industrial
revolution.
What characterizes our world today is the almost brutal intensity o f the race and the
fact that we are clearly losing it. The implications o f this trend are very disturbing to
contemplate. Our positions as world powers and our standards o f living are clearly at
risk. To reverse these trends we need to establish in our organizations an ongoing
focused process o f improvement. We need to disregard forever the myriad excuses that
have prevented us from facing the real problem.
We have been beaten at our own game. Our competition has worked smarter not
harder than us to gain their growing competitive edge. We believe this process o f
ongoing improvement can help restore our position. It is a faster, more economical,
more focused process than the Just-In-Time approach. If we use it, we can gain ground
in this race. Even so, as we rush to apply these processes we must learn from our
experiences.This process o f ongoing improvement is not the only or best way. We
must think even harder to find even better processes.
Good luck and much success in your efforts to win the race.

140
THERE IS NO FINISH LINE!

141
 Appendix A

A Process of
Ongoing Improvement
 A PROCESS OF ONGOING IMPROVEMENT
THE MARKETPLACE today is more crowded, faster-changing and more fiercely competitive
than at any time in history.
Industrial manufacturing is witnessing an intensification of the race for market-dominance:
the life-cycles of products arc shortening; zero-defects is becoming the goal of quality; new
machine technology is being introduced each year and systems to control production replace
each other at an unprecedented rate.
What was once relatively gradual change has in recent years turned into a race of exponentially
increasing intensity. Those unable to continually improve are falling behind, since success in
this environment requires more than a one-time improvement.
Each improvement docs of course buy some precious time, but the race in the market
continues relentlessly; the slope of the curve grows steeper and the time bought by any one
improvement becomes shorter.
Clearly, something far greater than a few sporadic improvements is now needed. Indeed,
the only way to secure and improve one’s competitive position today is by instituting a process
of ongoing improvement.
In the absence of such a process, the many improvements needed are certain to be sporadic
and fragmented wasting a great deal of energy, time and resources.
What is required is a process which will, at any moment, identify clearly the area where
an improvement will yield the maximum global impact.
This process must enable an organization to achieve the maximum gain from such improve­
ments, while simultaneously helping it to identify the area where the next improvement is
needed and to quantify the impact.
Since experience of how to implement such a process is both rare and badly needed we have
described below our understanding of the root cause of the inherent resistance to such a process
and some of ways we have developed for breaking down this resistance.

NOBODY would disagree that a process capable of generating an evolving, ongoing improve­
ment would be beneficial, but anyone who has tried to introduce any new process to an
organization knows all too well the multitude of obstacles encountered.
Experience shows that when these obstacles are closely examined, most will be found rooted
in the resistance of the people affected.
Although this resistance may take different forms in various environments, the fact remains
that in order to achieve the goal of introducing a process of ongoing improvement, one has
to accept and deal with the root cause of this reaction.
There is in fact nothing more difficult for any organization than change—any change. Each
improvement is by definition a change, although in the right direction, and each will predictably
be met with resistance, despite its beneficial potential.
As for a process of improvement, it is by its very nature a process of continual change. The
degree of resistance facing it may therefore be quite significant, although of course not all of
it conscious. Resistance may come from any tier of the corporate structure, for a change can
he as unwelcome in the boardroom as on the factory floor.
If, however, one accepts the existence of a fierce race to survive and flourish in the market,
and that one must have a competitive edge, then there is no alternative to instituting such a
process of continuing improvement.
Yet our deeply rooted, almost instinctive tendency to reject change makes the introduction
of such a process extremely difficult, if not impossible.
Change is opposed, from wherever it may come, not because the change itself was ill-
conceived, but simply because it is a change.
While one can try to crush opposition to change, this approach is a trying and time-consuming

144
effort even when successful. To yield to resistance means, however, to give up the only
effective, long-lasting remedy to the problem; hence the initial task is to change people's
attitude towards change, to neutralize opposition.
How can that be achieved?

Understanding Attitudes Towards Change

WE NEED first of all to tiy to understand the very different attitudes people have towards
improvements, as contrasted with their response to change.
Although both are by definition changes, “ improvement", unlike change, carries a positive
connotation. Is the key simply need to get others to understand the beneficial nature of the
change which we propose?
Let us look at our personal experience. How many times have we tried to explain an
innovation to others, whether they are our superiors, subordinates or peers, and felt that
somehow the message simply did not get through? We may even have provided flawlessly
logical arguments, crystal-clear explanations and eye-opening examples, but still our listeners
have remained unconvinced and stubbornly sceptical about the value of what we proposed.
We may often have been left with the feeling thaï, although they heard us, they didn't really
listen. Their energy seemed directed not at evaluating our proposal, but at finding out why it
would not work. If we succeeded at all in conveying our message and. perhaps in having it
embraced by all. then we managed to do so only thanks to great effort and perseverance.
Clearly, such a method of introducing an improvement is utterly inappropriate when trying
to gain acceptance of a process of ongoing improvement.
If we analyze our experience of introducing improvements, we are led inexorably to the
conclusion that the resistance we encounter derives more from emotion than logic. Our proposal
was perceived more as a change and less as an improvement and. as discussed before, change
invariably raises emotional resistance at every level.
In fact, emotions are rarely involved only at the receiving end. Our own emotions, not just
our logic, probably helped us to persist and, finally succeed in implementing our improvement.
If we are the champion of an idea, we tend to identify strongly with it. often so much so
that others regard it as our “ baby." The process of introducing improvements is partly an
emotional struggle, with the emotions involved in resisting change being finally overcome only
by the stronger emotions of the champion of change.
Our subject, however, is less the emotional nature of human beings than how to come to
terms with, and even possibly use. this emotional tendency so deeply ingrained in all of us.
If it were possible for everyone in an organization to convince themselves of the need to
adopt a process of ongoing improvement, then each individual would personally acquire an
“ ownership" of that idea. Then each person's emotional energy would be directed towards
embracing this process, rather than rejecting it.
But how can one hope to achieve such a sense of universal, yet individual ownership—and
have it occur simultaneously throughout an organization?

H5
 Attaining Individual Universal Ownership
INDIVIDUAL, yet universal, ownership— it sounds impossible, a contradiction in terms, but
that is precisely what we need to achieve.
If we wish to create a climate in which a process of ongoing change will be welcome
throughout an organization, we need to elicit such a widespread commitment of that individuals
there is a universal acceptance of the process.
If one were to build a tool to perform this miracle, what should characterize it?
It probably should be inexpensive enough to permit mass distribution, be readily available,
require no set-ups or supporting resources, be convenient to use and be capable of being
adapted by an individual to his pace or preferences.
Since the individual is our focus maybe a book would be an effective vechicle.
However, not every kind of book can become a tool for acquiring personal ownership of
an idea. Textbooks, though they convey information, usually fail to excite the imagination.
Readers may absorb the words, but they seldom take ownership of the ideas. For this to happen,
a reader needs a book that involves him personally and if possible mirrors his own experience.
Because the ancients appreciated this idea, analogies have been used throughout the ages
as a means of teaching and communicating ideas. When this approach is well done in books,
plays, movies and the like, the reader or viewer identifies with a character to such a degree
that he mentally decides how the character should try to solve his problems.
We conceived the idea of writing The Goal, which many of you undoubtedly have already
read, by taking this teaching tool— an analogy—and embodying it in a novel. A novel makes
the acquisition of information enjoyable and encourages readers to identify with the situation
and characters employed.
We knew that The Goal had to portray the industrial world vividly and realistically in order
to be effective. It needed faithfully to reflect the pressures, problems and daily decisions
confronting an individual in that environment. Moreover, it ought ideally to cause the reader
to identify with the conflicts faced by the protagonist, Alex Rogo and allow the reader to share
Rogo’s anxiety to find solutions and save his plant.
We knew the The Goal would not be effective in creating ownership of the ideas unless it
drew the reader into the story and led him to an understanding of the cause-and-effect rela­
tionships between the problems and the solutions in the analogy.
We hoped the reader would-be struck by parallels to his own environment and see a spectrum
of problems, many of them personally experienced. We wished him to experience vicariously
the struggle to develop a commonsense approach to a class of problems leading to the first
glimmer of light—a general solution.
As more specific implementable solutions are derived from the general solution, the reader
would, we hoped, begin to perceive the pattern of how the specific solutions are generated.
Sooner or later, we believed, that reader would make the intuitive leap from the thought process
described in the analogy to applying this process to the problems encountered in his own
environment.
He would then have had what psychologists call an “ Aha! experience” — and the solutions
he would then invent whether they are to Rogo’s problems or those of his company would be
his own.
However, we believed that, no matter how many discrete solutions are implemented, they
will not add up to a process of ongoing improvement.
We wanted the reader to develop a sense of ownership, not just of a specific “ invention” ,
but also of something much more nebulous. We wished him to take ownership of the idea of
instituting a process of ongoing focused improvement in his own organization.
What is meant by such a process of ongoing focused improvement? In any organization a
very small number of constraints govern the overall level of performance. If these few con-

146
straints can be relieved, the performance of the entire organization will be raised significantly.
Our experience has shown that other improvements which do not involve relieving primary
constraints may have a positive impact, their effect is of a much lower order of magnitude.
The first step in a process of ongoing improvement, then, is to identify precisely the primary
constraints and to focus all our efforts on alleviating them. Once this has been achieved, the
entire organization will be raised to a new level of performance which will now be limited by
the existence of some new constraints.
It is the very essence of a process of ongoing improvement that we cannot allow ourselves
to be satisfied with a higher performance level, but must be motivated to focus, as rapidly as
possible, on those emerging constraints which limit our attainment of an even higher perform­
ance for the entire organization.
This is a never-ending process. Constraints may shift from one sector of an organization to
another and may even shift to the market outside the organization itself.
Nevertheless, regardless of where they originated, it is always possible to take actions within
the organization to alleviate the major constraints.
We wanted to portray this process to the extent that the reader could develop the feeling of
personal ownership that such a process could be and, moreover, must be, possible within his
own organization.
We also believed that the book must portray this process vividly and that it must describe
several cycles in the process. But even that was not enough. Each of the cycles had to describe
a situation where the constraints were so common that many readers would have already
encountered a similar constraints.
If the book succeeded in depicting several cycles that the reader could relate to in his own
environment, then the chances would be quite good that the reader would identify with the
entire process of ongoing improvement.
The book’s requirements, we knew, were extremely demanding. But we also knew that, if
we did not succeed in meeting all of them, then we would probably fail to overcome the
inherent resistance to a process of ongoing improvement.
If your are like the readers of The Goal who have contacted us, you may well feel a strong
urge after reading it to do at your plant what Alex Rogo did at his.
You may well have become convinced that the types of changes made by Alex should also
be implemented in your plant. You may also believe that such changes, while extremely
beneficial, are not sufficient in todays world it is onlly by instituting a process of ongoing
improvement can your plant, like Alex’s survive and flourish.
You may right now be trying to address some of the most persistent problems of your
organization, using the same approach described in the book.
We say this because it is a process undergone by many of the readers of The Goal. Of
course, the closer a reader is to the particular environment depicted in the book, the stronger
his sense of identification with it. Although the principles of the solution may suit many fields,
such as hositals, governments services as well as industry, it is easier for people from man­
ufacturing to find similarities to their environment and to identify with the people and situations
depicted in The Goal. People coming from production and assembly plants tend to identify
with Rogo’s situation even more than those from process plants.
Those readers who have communicated with us tell us that the scenario of the book is true
to life, that the anxiety felt by Al Rogo, whether caused by his boss, his people or even his
family, is one with which they are all too familiar. Many have accused us of sneaking into
their plants and homes and writing The Goal about their own problems.
They have found The Goal appealing not only because it realistically describes the events
accompanying the introduction of a process of ongoing improvement, but also because it

147
highlights the burning issues and concerns of contemporary manufacturing.
By being exposed to the experience a particular plant had in introducing a particular process
of ongoing improvement, they have come to see how such a process works.
They have seen how the performance of an entire plant is influenced by the identification
and eventual opening up of a bottleneck. Moreover, they have seen how further improvement
is blocked by yet another problem—inventory management.
Their reading has shown them that the solution to this problem and, as a matter of fact, to
any problem, creates a new reality in the plant which gives rise to still more problems. And
that new, external, obstacles emerged as these internal problems were resolved.
Rogo discovered that, once his plant's immediate manufacturing problems were overcome
and performance improved markedly that marketing became his biggest problem.
Nevertheless, Rogo was able to deal with this external problem through internal actions. By
shortening his lead times he was able to get additional orders.
What readers come to appreciate is that the reality of the plant is always shifting—that every
improvement, as it produces benefits, also creates a new reality which must be addressed,
resolved and changed again. It is this process which is never-ending.

THE hundreds of letters we have received from readers of The Goal show that most of them
feel a compelling urge to start changing things at their plants.
But a word of caution is required at this point. Whether an improvement deals with the
tooling of a single machine or with the marketing policy of an entire plant, it remains an
isolated, local action resulting from the initiative of only certain individuals. However bene­
ficial, these improvements are not enough.
What is needed goes beyond the isolated, uncoordinated efforts of individuals trying to do
their best. It calls for the unified efforts of an organized group. Only that can give the
implementation of a process of ongoing improvement a fighting chance.
We are therefore faced with the problem of how to bring an entire group to view the process
of ongoing improvement in the same light and to embrace it together.
Our experience is that even after people have reached general agreement on the major issues,
they tend to spend incredible amounts of time and effort arguing about the specific procedures
for resolving these issues.
How, then, does one resolve this problem?

Reaching a group consensus

EARLIER, we dealt with the problem of how an individual can acquire ownership of the idea
that a process of ongoing improvement needs to be implemented in his organization.
To achieve it, we used a book containing very detailed and explicit examples. Now we are
confronted with the opposite problem, how to turn the ownership of many individuals into a
group consensus.
Addressing a group with arguments based on a specific, detailed example is actually the
best way of ensuring that consensus will not be attained. Everyone in the group will tend to
draw differing parallels to his real-life situation and the only likely result is a fruitless and
heated debate.
We need instead to base a process of ongoing improvement upon a very precise and logically
flawless derivation of general rules and procedures.
One tends, however, to regard such general rules and procedures as vague and subject to
more than one interpretation. Trying to encourage ownership through a group process of
formulating such general rules and procedures would most probably result in something vague
to the point of meaninglessness.

148
 Indeed, can general yet powerful approaches applying to most industrial organizations be
formulated?
The Goal contains some such rules but, though they are useful in focusing on and relieving
major constraints, they do not address the process of ongoing improvement itself.
The example in The Goal merely suggested such a process. If, however, we don’t succeed
in formulating the general rules and procedures for such a process, the probability of achieving
consensus is extremely low.
Without such consensus the seeds of resistance will reappear. As the organization strives
to develop and institute its own process of ongoing improvement, these seeds will grow into
pockets of fierce resistance.

On the other hand merely presenting the appropriate rules and procedures to a group will
not ensure their acceptance. Such a presentation needs to include the entire step-by-step de­
rivation of this approach.
Consensus will be reached only if this derivation starts from a generally agreed-upon picture
of the situation in industry today and proceeds using very precise, well-defined arguments,
making sure that no gaps or even perceived flaws leave an opening for misunderstanding. The
logic must be so strong that it is perceived as common sense. This is what we have tried to
accomplish in The Race.
But reaching consensus will not by itself yield the desired result. Another step needs to be
undertaken.
It is a far from trivial task to make the transition from general rules and procedures into a
workable process of improvement. The next difficulty encountered is typically the question
of applicability.
The approach described in The Goal and The Race has three different avenues of applicability,
the first being the managerial sphere.
Managements today often tend to perceive problems and effects as independent events. One
of the most appealing aspects of this approach stems from coming to a realization that many
seemingly unrelated events in plant arise from a very few common sources.
The perception of these cause-and-effect relationships is to a large extent new and certainly
provides a much more powerful way to run a business.
The second area of applicability is the logistical one. The drum-buffer-rope method used
in The Goal and more precisely analyzed and defined in The Race is a very attractive logistical
system.
Its appeal comes from the fact that it eliminates both the confusion and the expediting mode
in which most plants find themselves.
Buffer-management brings the local improvement activities under the same umbrella. In
other words, the logistical applicability reduces confusion and brings order to other daily
activities.
The third area of applicability is the behavioral one. As can be deduced from The Goal and
as is further highlighted in in The Race, a process of ongoing improvement has significant
behavioral ramifications on the shop floor, as well as within top management.
Applicability in these three areas has a significant impact on the bottom line.
In evaluating this impact, wc should not ignore the strong connections and interrelationships
among the three. Difficult as it might he, we must evaluate the degree of applicability in order
that the consensus does not turn rapidly into frustration.
The question of exactly how to best accomplish this and overcome many of the other
obstacles to rapidly instituting an effective process.of ongoing improvement is beyond the
scope of this discussion.
 We know that The Goal and The Race are very helpful in increasing individual ownership
and group concensus of some of the changes needed in our plants. They have also had an
impact on creating an awareness of the need for a process of ongoing improvement and how
such a process might work.
They are by no means sufficient. Considering the precarious position of Western industry
and the small window of time remaining much more must be done. New, innovative tools for
creating a widespread ownership of the changes required by all functions is needed. General
rules and procedures like the drum-buffer-rope approach are vitally needed in almost all areas
of manufacturing.
This is the real challenge facing Western industry. We hope that this sharing of our perception
of the need for a process of ongoing improvement and the root causes of the resistance to it
coupled with our efforts to develop such a process are helpful in this effort.

ISO
 Appendix B

Can You Win

at Managing
The Production Game

151
152
INTRODUCTION
These quizzes illustrate five different types of decisions we
face daily in our businesses. They are much less complex than
the situations you must analyze, but the principles required to
solve both the quizzes and your real life problems are the same.
People derive great value from working with the quizzes
before they turn to the answer book. We encourage you to do
the same. W e seem to learn more from an active effort to solve
a problem rather than passively being shown how to solve it.
Good luck with the quizzes. Look forward to learning of your
experiences.

Robert E. Fox

153
A CAUSE OF EXCESS
INVENTORY?
Two workers are engaged in
production of a product and its
spare part. The resources
required, rates of production,
operational steps, and market
potentials (demands) are given in
the diagram. There is only one
Worker V and one Worker W per
shift. The plant operates 24
hours a day (3 shifts), 5 days a
week. The workers cannot
substitute for each other and
there is no other work except as
detailed in the diagram.

Legend
Resource (e g Worker V)
Production Rate (e g. 4 min unit)
Operation Number (e g A-30)

154
• Schedule the two types of workers for a typical day (24 hours).

» Estimate inventory accumulation (in units) after one day, after one
week, and after one month according to your one-day schedule.

Buildup of Excess Inventory

Day 1 Week 1 Month 1

Operation A -10

Operation A -20

Operation A -30

Operation B-10

Operation C-10

• Are there any steps you might take to lower the level of inventory
accumulation without losing sales?

• In light of the above, draw some conclusions about the following

ideas:

— Worker "piece incentive" as a reward system

— Worker efficiency as local performance measurement

155
HOW FREQUENTLY SHOULD
YOU SET-UP?
The market potential (demand) for
Product A and Product B exceeds
the plant's capacity. There is only
one Worker U, one Worker V, and
one Worker W per shift. The
plant operates 24 hours a day (3
shifts), 5 days a week. The
workers cannot substitute for
each other and there is no other
work, except as detailed in the
diagram specifying the required
resources (type of worker), rates
of production, and sequence of
operations.
A set-up of 180 minutes is
required of Worker W to start
production and whenever he
switches from one product to
another. You may assume that no
scrap occurs in this set-up.

Legend
Resource (e.g. Worker U)
Production Rate (e g 20 min /unit)
Operation Number (e g A-20)

156
 What is the most efficient set up frequency that Worker W should
follow? Express the result in terms of hatch si/e

• If you feel you lack certain information essential to providing a

meaningful answer to this problem (like worker salary or carrying
cost of inventory), please outline the types of additional information
you feel are required.

157
DOES COST DRIVE YOUR
MARKETING?
Two different resources (Workers
V and W) are engaged in the
production of 4 products. The
diagram specifies the required
resources and rates of
production. It also shows the raw
material costs and product selling
prices. The plant operates 24
hours a day (3 shifts), 5 days a
week. There is only one Worker V
and one Worker W on each shift
and they cannot substitute for
each other.
At the given prices, the market is
willing to buy any quantity that
the plant is able to produce,
provided that the plant will offer
the total spectrum of products.
This constrains the product mix
that can be sold, such that no
single product can exceed the
others by a factor of more than 10.
(e g. if Product A is to be sold
in quantities of 1,000/day, the
plant must produce at least
1 0 0 /d a y of products B, C, and D
as well. The total operating
expenses of the plant —
excluding the purchase of
materials — are $12,000 per
month (22 working days).

• What mix of the four products will you choose to offer the market?

Product A B C D

Ratio

• According to the quiz data and your choice of product mix. what
will be the net profit of the plant in a typical month?

158
 Products
A B C D

Legend
Resource (e g. Worker V)
Production Rate (e.g. 5 min./unit)

Raw Materials Costs

Selling Price

159
HOW SHOULD WE PLAN A
JOB?
An order for 1,000 units was
received by your plant. The
diagram specifies resources Sales Price
needed, production rates, and the $450 per unit
sequence of operations. The time
to set-up each resource is 10
hours. Notice that Resource V is
used for two different operations.
Machines U,V,W, and Y are
dedicated to this order, along
with a crew of four workers per
shift. The plant operates 24
hours a day (3 shifts), 5 days per
week. There is only 1 machine of
each type. Each machine is
activated by 1 worker who is also
capable of setting it up.

Legend
Resource (e y Machine Y)
Production Raie (e y 100 min unit)
Opérai ion Number (e y A 50)

160
 Draw a schedule for the job.

Assume that the operating Likewise, assume that payment

expenses of the plant — for the
designed resources and
manpower, including overhead —
are $7,500 per week and are
paid at the end of the week.
for material is made when raw
material is actually used at the
first operation (A-10). Finally,
money is received whenever a
finished unit is completed.

• Plot the cash flow that results from this schedule:

200 $ (in thousands)

100

Months
0
8 9 10
100

200

300

400

• II the plant h.is only $ 2 00,000 in cash, should it accept this order?

161
SHOULD CPM/PERT BE
USED FOR PROJECT
PLANNING?
The project has to be completed
in 150 working days from today.
To complete the project, all 3
tasks (A,B,C) have to be
completed. The diagram specifies
the resources (teams), operation
times, and the sequence of
operations for the project. There
is only one Team V and one
Team W available for the project.
The lead time for procurement of
the materials needed (described
in the diagram as circles) is 50
working days for each material.

Legend
Resource (e g Team W)
Operation Time (e g 45 days)
Operation Number (e g A-20)

Raw Materials

162
 Use CRITICAL PATH or PERT to determine when to order each
material.

Materials for Task Day to Order

(in working days from start)

A
B
C

• Draw the schedule for the project (Gantt Chart).

Team V

Team W
--------------------------- ---------------------------- ---------------------------- ---------------------------- — ►
50 100 150 200 days

• Will you finish the project on time?

• Try again — using your intuition this time. When is the earliest
possible time you can finish the project?

Team V

Team W
--------------------------- ---------------------------- ---------------------------- ------------- ■ — ►
50 100 150 200 days

163
164
 Appendix C

How to Win
at Managing
The Production Game

165
166
INTRODUCTION
These Quizzes are not a measure of intelligence. They either
are solved quickly with little effort or require hours and much work.
It’s not how smart you are but how you approach the problem. If
we rely on our intuition, they are relatively simple. When we use
what we Ve been taught, they are often very difficult.
Does this suggest that to be successful in manufacturing, we
must first unlearn things? The answer unfortunately is yes. First
and foremost, we must change our perception of “cost”. Cost is
deeply imbedded in how we typically make manufacturing deci­
sions as illustrated by the conventional rules shown on page
12. When we use these conventional rules (our perception of
cost) to decide how to use labor efficiently, determine EOQ’s
and to calculate product cost, we get the wrong answers. Not
only in these quizzes, but also in our businesses.
Secondly, we need to recognize that MRP should not be used
for planning and scheduling. The last quiz on PERT/CPM shows
the fallacy of back scheduling and ignoring capacity (MRP).
MRP is an excellent tool for collecting, organizing and
disseminating data and information. Let’s use it for that pur­
pose. It does not and will never (unless you have infinite capaci­
ty) result in realistic, stable and profitable schedules.
Changing our perception of how cost and MRP should be us­
ed is an essential step to correct answers and greater pro­
fitability.
When we apply only our intuition, we get the correct answers
to these quizzes much more easily. We also make better and
more profitable decisions in our businesses. You will see your
intuition mirrored in the global principles on page 13. When we
formalize our intuition into rules like these global principles, then
solving these quizzes and making business decisions is even
easier and more profitable.
Facing up to the challenge of unlearning before we can move
forward has been both humbling and exciting for all of us. It all
depends on how you decide to look at it. You might be in­
terested to know that people with limited understanding of how
we typically make manufacturing decisions get to the correct
answers more quickly. They seem to rely more on their intuition
rather than reverting to what we have learned about how things
should be done — the conventional rules.
We hope these quizzes and T h e G o a l assist you in this
journey. Our experience is that it’s a very exciting and rewarding
one. Good luck!

R o b e r t E. F o x

167
A CAUSE OF EXCESS
INVENTORY?
The usual concern in
constructing a schedule is to
make sure that each worker is
scheduled to be kept busy all the
time. A global approach is s
different. The first step is to
identify the constraints inherent
in the problem. A brief
examination of our case shows
that Worker W is a constraint
since he cannot produce more
that 144 items a day (24 x
6 0 /1 0 ) limiting the sales of the
product to that number. The
other constraint is an external
Typical Daily Schedule
Notice that Worker V is scheduled
not to produce anything for 3.13
hours/day. This is an application
of Rule: 3
"U TILIZA TIO N A N D
A C TIV A TIO N OF A
RESOURCE ARE NOT
S Y N O N Y M O U S ".
168
one, the market potential for the
spare part.
The schedule of Worker V is
arranged to satisfy the constraint
to produce 144 units/day at
operations A-10, A-30 and C-10
and 244 units/day (144 + 100) at
operation B-10. Any production
above these quantities will not
yield more sales - just more
unneeded inventory. Thus a
typical schedule for the two
workers will look as follows:
By obeying such a schedule,
accumulation of excess inventory
is avoided. Trying to activate a
worker to his full potential will
not result in additional sales but
only in the accumulation of
inventory before A -20 (due to the
limited capacity of Worker W ) and
after B-10 (due to the market
restriction and due to the
limitation of type-A parts
available for assembly).
If we assume the common spare time in overproducing at
practice is followed of activating A -10 and B-10, the buildup of
V to its full potential, splitting its excess inventory will be:

Buildup of Excess Inventory

Day 1 Week 1 Month 1

Operation A-10 47 235 1034

Operation A-20 — — —

Operation A-30 — — —

Operation B-10 94 470 2068

Operation C-10 — — —

Notice that the traditional plant performance. In our

performance measurements of a example, holding about 50 pieces
worker, whether piece incentive (60 x 8 /1 0 ) after A -10 and 80
or efficiency measurements, pieces (50 + 30) after B-10 will
encourage the scheduler, the give sufficient protection to the
foreman and the worker to use plant to overcome deviations of V
the worker's potential for of up to a full shift of production.
producing to the maximum. If we
want to avoid the buildup of
excess inventory, we must
replace such performance
measurements. A logical
replacement is "adherance-to-
the-schedule". Workers whose
capacity does not constrain the
performance of the plant should
produce according to a reliable
predetermined schedule, and be
encouraged not to overproduce
as well as not to underproduce

Adequate stable levels of

inventories should be maintained
to insure that deviations from the
schedule will not jeopardize the

169
HOW FREQUENTLY SHOULD
YOU SET-UP?
The usual way of setting the
frequency of setups is to
minimize the cost/unit (the sum
of the setup cost plus the
carrying cost). Many articles have
been published dealing with this
problem, under the title
"Economic Batch Quantity". In
most places a rule of thumb is
used, setting the production time
to be larger than the setup time
by some factor (usually between
4 to 10 times).
The conventional attempts to
save setups on non-bottlenecks
do not really save anything. They
simply increase the amount of
unneeded spare time. This is a
demonstration of Rule 5:
"A N HOUR SAVED AT
A N O N -B O TTLE N E C K
IS J U S T A M IR A G E "

A global approach starts by

identifying the constraints of the
system. In our case, U and V are
clearly the constraints since each
one of them requires
20 m in/unit-pair, while W
requires only 10 m in/unit-pair
(5+5). Thus W can be
productively utilized for producing
parts only 50% of its time. Any
production beyond that will just
yield inventory but not additional
sales. Since 50% of W 's time✓ is
free, why not utilize it for setup,
thus reducing inventory without
any reduction of sales?

Since the setup time at Worker W

is 3 hours, the batch size at each
of its operations should also be 3
hours, or 36 parts (3 x 6 0 /5 ).
However, we should avoid full
utilization of W to prevent any
fluctuation of W, V or U from
causing lost sales. Therefore, the
most efficient batch size would
be a little larger than 36. This
will result in sufficient idle time
and inventory to cushion
fluctuations

170
DOES COST DRIVE YOUR
MARKETING?
The conventional cost accounting
approach to determining the
cost/unit for each product would
lead to choosing A and C as most
profitable products. This is
because for both A and C, the
"price minus raw material" cost
is $12 (30 - 3 - 5 - 10) whereas
C o n v e n tio n a l C o s t in g
Product A B
Ratio 10
A global approach uses a
different way of allocating
indirect costs - according to the
time each product spends on the
bottleneck. This is consistent
with Rule 4:
"A N HOUR LOST A T A
BOTTLENECK IS A N HOUR
LOST FOR THE ENTIRE
SYSTEM "
This means that the output of the
entire system is linked directly to
the time available on the
bottleneck. This points up the
Global Costing
Product A B
Ratio 10
for B and D it is $10 (32 -7 - 5 - 1 0 ) .
Note that the indirect costs of
$ 1 2 ,0 0 0 /month are allocated
equally to all four products since
the labor time is the same for
each. Therefore, the best mix
would appear to be:
C D
1 10 1
opportunity for improving
revenues by changing the mix to
favor products requiring relatively
less bottleneck time.
In our case, A and B are more
favoured than C and D because
the time on the bottleneck W is
20 m in/part which is only 2 /3 of
the bottleneck time needed for C
and D. Furthermore, A has a
more favorable "price minus raw-
material cost" ($12) than B ($10),
as already noted. Thus, the best
mix would be:
C D
1 1 1
171
Note that for our case, constraint
W remains the bottleneck
regardless of product mix,
because it is the bottleneck for
each product individually.

The net profit may be calculated

and compared for the
conventional versus the global
approach. Choosing the mix
suggested by the conventional
approach, the limitation of W will
allow for the monthly processing
of 576 units each of products A
and C and 58 units each of
products B and D This mix leads
to a net profit of $2984. By
comparison the OPT mix will be
1131 units for product A and 113
units each for products B, C and
D. The OPT mix gives a net profit
almost two times larger of $5188.

It should be noted that in a real

life situation the attainable
increase in the net profit is often
even larger because the time
spread encountered between
different products on the
bottleneck is usually even larger
than for the case here.

172
HOW SHOULD WE PLAN A
JOB?
The conventional approach is to
treat each order as a single batch
to be done in full at each
successive operation. In our case.
C o n v e n tio n a l S c h e d u le
A -50
A -40
A -30
A -20
A -10
1000 2000
A calculation of the cash flow
shows that the money being
spent for the first 3 708 hours
reaches a total cash outflow of
$382,500 before the sales
revenue begins to reverse the
flow. By the end of the period,
the job has resulted in a net loss
of $37,500. (not taking account
of the additional loss due to the
interest cost on the money
required).
A global approach starts by
identifying the constraints.
Machines U. W and Y are the
bottlenecks which must be fully
activated. Machine V is a
non-bottleneck so we can utilize
an order for 1000 would take
5385 hours as shown on the
schedule:
300 0 4000 5000 6000 hours
its spare time very effectively to
decrease the batch size with no
penalty for the additional setups.
At 15 units/batch, V would be
fully utilized but to avoid loss of
throughput due to fluctuations in
U, V, W and Y, we should ideally
choose a somewhat larger batch
size. All major benefits will be
obtained even with a batch of 30.
The result is a schedule with one
worker on V alternating
frequently between operations
A -2 0 and A -40 and with the other
three workers dedicated to U, W
and Y, all working in parallel. A
sample of the schedule is shown
here.
173
Startup of OPT Schedule
The overlapping of operations
A - 10, A -30 and A -5 0 is made
possible by moving small transfer
batches through the system
while maintaining large process
batches only at the constraining
resources. The net effect is a
drastic reduction in the length of
the project from 5308 to 1807
hours. The $ 3 7,5 0 0 loss is
Cash Flow
174
converted to a net profit of
$187,500. The competitive edge
of the reduced cycle time should
have an important effect on
future sales. Also, the initial
negative cash flow is reduced
from almost $ 4 0 0 ,0 0 0 to just
$20,000, as shown in the
diagram.
A very important side effect of
reducing the cycle time and
inventory is the effect on quality
Consider the effect of a defect in
the first operation A -10 which is
not noticeable until the last
operation A-50. Under the
conventional approach, the defect
is detected months after it
occurred and after the whole A - 10
process has been completed
and disassembled. W hat can be
done to correct process A -10
after the discovery is too little
and too late.

By contrast, the global schedule

will lead to the discovery of the
defect after a few days when
only a tiny fraction of the parts at
the A -10 operation have been
processed. The identification and
quick correction of the defect in
the A -10 operations can be
crucial.

This case illustrates how

important it can be to follow.
Rule 7 which states:

"THE TRANSFER BATCH M A Y

NOT A N D OFTEN S H O U L D
NOT BE EQ U AL TO THE
PROCESS B ATCH ".

175
SHOULD CPM/PERT BE
USED FOR PROJECT
PLANNING?
The Critical Path Method (CPM)
or PERT first determines and
labels the longest path in time as
the critical path. The schedule for
the critical path is obtained by
backing off from the final due
date. Other parallel paths have
slack time relative to the critical
path resulting in the option of an
"early" or "late" starting date,
relative to the critical path.
In our case, the critical path is
task A (65 days) and tasks B and
C have a slack of 10 and 15 days,
respectively. Thus, backing off
from the due date of 150 days,
the order point for A is 35 days
CPM/PERT Schedule
-------- £ --------------------- “ 8 ---------C ----------------
0 50 100
This is not an unusual
experience. A project may appear
at the planning stage to have
plenty of time before the raw
material needs to be ordered and
only after the project gets
underway do we find that even
going to extensive overtime is
insufficient to bring the project to
completion on schedule.
176
(150 - 5 - 60 - 50), for B is 45
days (35 + 10) and for C is 50
days (35 + 15).
Even if the vendor is reliable and
the raw material arrives as
scheduled, we will find that the
project is extending much longer
than anticipated. An examination
of the Gantt chart, which obeys
the priority scheme of
CPM/PERT, for teams V and W
shows that the competition for
resources is responsible for
pushing the project completion
date to 245 days, way beyond the
required 150 days.
i 1
i
» p
| Ai I o L
150 200 245 days
Let us now plan an alternative
schedule, replacing the critical
path concept with our own
intuition and keeping in mind the
constraints on the resource
teams. By sequencing team V
differently, the project can be
greatly shortened as shown in
the Gantt chart:
Global Schedule

n A
Team V
i

Team W
! c B 1A

50 100 150 200

days

The results are a reduction of This case is an illustration of the

project completion from 245 days
to 145 days (which is within the
required time). The average time
the inventory is held is reduced
from 245 - (3 5 + 4 5 + 5 0 )/3 = 202
days to 145 -(0 + 5 + 3 0 )/3 = 133
days.
Note that when we compare to
the previous sequencing (PERT),
we see that non-bottleneck W
was converted into a constraint
which greatly lengthened the
project. This phenonema of the
shifting or "wandering" of
bottlenecks and the resulting
waves of "work - no work" is
often observed in real life. W e
see here why such phenomena
occur - because of improper
sequencing of tasks on the
bottleneck.
importance of Rule 9 which states:
"S C H ED U LE S S H O U L D BE
ESTA BLISH ED BY LO O KING
AT ALL OF THE
C O N STR AINTS
S IM U L T A N E O U S L Y . LEAD
T IM E S ARE THE RESULT OF
A SCH EDU LE A N D C A N N O T
BE P R E D E T E R M IN E D ."
As difficult as it may be to
establish priorities for proper
sequencing, it is nevertheless
essential for getting the
performance the system is
capable of.

It is worth recalling that PERT

methods are the basis for all
MRP systems. Clearly, it is
dangerous to ignore the resource
constraints at the initial stage of
planning, even if these
constraints are afterwards loaded
in and massaged to give the best
possible schedule that may be
generated from that inaccurate
starting point.

177
WHAT RULES DRIVE
CONVENTIONAL RULES
• Balance capacity, then try to maintain flow.

• Level of utilization of any worker is determined by its

own potential.

• Utilization and activation of workers are the same.

• An hour lost at a bottleneck is just an hour lost at

that resource.

• An hour saved at a non-bottleneck is an hour saved

at that resource.

• Bottlenecks temporarily limit throughput but have lit­

tle impact on inventories.

• Splitting and overlapping of batches should be

discouraged.

• The process batch should be constant both in time

and along its route.

• Schedules should be determined by sequentially:

• Predetermining the batch size.
• Calculating lead time.
• Assigning priorities, setting schedules according
to lead time.
• Adjusting the schedules according to apparent
capacity constraints by repeating the above 3
steps.

MOTTO
The only way to reach
a global optimum is by
insuring local optimums.

178
YOUR BUSINESS?
GLOBAL RULES
1. Balance flow not capacity

2. The level of utilization of a non-bottleneck is not

determined by its own potential but by some other
constraint in the system.

3. Utilization and activation of a resource are not

synonomous.

4. An hour lost at a bottleneck is an hour lost for the

total system.

5. An hour saved at a non-bottleneck is just a mirage.

6. Bottlenecks govern both throughput and inventories.

7. The transfer batch may not and many times should

not be equal Io the process batch.

8. The process batch should be variable not fixed.

9. Schedules should be established by looking at all of

the constraints simultaneously. Lead times are the
result of a schedule and cannot be predetermined.

MOTTO
The sum of the local optimums
is not equal to
the global optimum.

179
$ 15.00

On the surface, Eli Goldratt and Bob Fox are a marked contrast in background and styles.
Goldratt was born and raised in Israel during a tumultuous period in the Middle East and has
fought in three wars. He left a promising academic life after receiving a doctorate in physics
from Bar llan University to launch a seemingly Don Quixote venture. Even though he never
worked in a plant or even visited one in the U.S., Goldratt’s goal was to teach Western
manufacturers how to schedule their operations.
Appearing in an open-collared white shirt without coat or tie, Goldratt is often described
as brash, evangelistic and a killer of sacred cows. Nevertheless, he has in seven short years
become an internationally recognized thought leader in manufacturing. His OPT logistical
system (small OPT) and the OPT process of ongoing improvement (big OPT) are revolutionizing
how manufacturing is conducted. Goldratt’s developments have been adopted by more than
100 major manufacturers and are already being taught at more than 50 colleges and universities
in the Western world. Some schools are even devoting entire courses to his inventions. His
first book, THE GOAL, a novel of business, has become an underground best seller in board
rooms, colleges and the factory floor and is already available in eight languages.
Bob Fox’s background was much more conventional starting in the security of a small town
in the heartland of Nebraska. After studying engineering and business at Notre Dame and
Carnegie Mellon Universities, Bob began his industrial career as a foreman. Progressing rapidly
through a series of production and engineering positions, he served as a Vice President of
Manufacturing. After a stint as General Manager, he returned to and became a partner at Booz,
Allen and Hamilton, the internationally recognized management consulting firm. An articulate,
polished, professional speaker, he lectured frequently before professional and industry groups.
The emergence of Japanese industrial competition caused Goldratt’s and Fox’s paths to first
cross, and eventually linked their energies to find better solutions for Western manufacturers.
Fox was one of the first to publicly describe the techniques and success of Japanese manu­
facturers. He recognized that Goldratt’s computerized OPT logistical system, although devel­
oped totally independently, was remarkedly similar to the manual Japanese Kanban approach.
Joining forces with Goldratt and other colleagues at Creative Output, they have successfully
challenged such manufacturing sacred cows as the use of efficiencies, variances and standard
costs. More recently, they have expanded OPT into a process of ongoing improvement that
many believe is superior to the Japanese Just-in-Time approach.
On the surface, Goldratt, and Fox’s similarities seem to be confined to a love of fine cigars,
but they share an intense drive and deep conviction that the growing dangerous imbalance of
manufacturing capabilities between the West and East must and can be changed. In conjunction
with a number of major companies and universities, they are now building the Abraham Y.
Goldratt Institute to develop the know-how to change this imbalance.